JP2017534421A - Rapid detection of bleeding before, during, and after resuscitation by infusion - Google Patents

Abstract

In some cases, new tools and techniques are provided for assessing, predicting, and / or estimating the probability that a patient will bleed non-invasively. In various embodiments, tools and techniques are provided for quickly detecting bleeding before, during and after infusion resuscitation, possibly in real time.

Description

Cross-reference of related applications
[0001] This application is filed under United States Patent Act 119 (e) with the following co-provisional application: US Provisional Patent Application No. 62/064 by Mulligan et al., Filed Oct. 16, 2014. 809, “Rapid Detection of Breeding Fluid Fluid Resuscitation” (Attorney Docket No. 0463.14PR, hereinafter referred to as “'809 Application”), and US Provisional by Mulligan et al., Filed Oct. 16, 2014. Claim the benefit of Patent Application No. 62 / 064,816 “Assessing the Effect of CPR” (Attorney Docket No. 0463.15PR, hereinafter referred to as “'816 Application”), both incorporated herein by reference It is. This application is related to co-pending US patent application _________ “Assessing the Effect of CPR” (Attorney Docket No. 0463.15) filed on the same day by Mulligan et al. Claim priority to the 809 and '816 applications.

  [0002] This application is based on the following application: US Provisional Patent Application No. 61 / 904,436 “Nonvasive Monitoring for Fluid Resuscitation” filed on 14 November 2013 by Mulligan et al. No. 0463.11PR, hereinafter referred to as “'436 application”), filed Nov. 14, 2014, filed Nov. 14, 2014, US patent application Ser. No. 14 / 542,423, Noninvasive, by Mulligan et al. "Monitoring for Fluid Resuscitation" (Attorney Docket No. 0463.11, hereinafter referred to as "'423 application"), and US Provisional Patent Application No. 61 by Mulligan et al., Filed Nov. 18, 2013. By Mulligan et al., Filed on Nov. 14, 2014, claiming priority to No. 905,727 “Noninvasive Hydration Monitoring” (Attorney Docket No. 0463.12PR, hereinafter referred to as “727 Application”) No. 14 / 542,426, “Nonvasive Hydration Monitoring” (Attorney Docket No. 0463.12, hereinafter referred to as “'426 application”), all of which are hereby incorporated by reference. Embedded in the book.

  [0003] This application is also filed on November 6, 2013 by Mulligan et al., US Provisional Patent Application No. 61 / 900,980 “Nonactive Predictive and / or Estimative Blood Pressure Monitoring” (Attorney Docket Number). No. 14 / 535,171, “Nonvasive Predictive,” filed Nov. 6, 2014, claiming priority to US Pat. No. 14 / 535,171, filed Nov. 6, 2014, claiming priority to 0463.10PR (hereinafter “the '980 application”). and / or Estimative Blood Pressure Monitoring ”(Attorney Docket No. 0463.10, hereinafter referred to as“ '171 Application ”), both of which are referenced More are incorporated herein.

  [0004] This application is also filed on July 22, 2011, by Grudic et al., US Provisional Patent Application No. 61 / 510,792, “Cardiovascular Reserve Monitor” (Attorney Docket No. 0463.05PR, hereinafter). US Provisional Patent Application No. 61 / 614,426, “Hemodynamic Reserve Monitor and Hemolysis Control” (Attorney Docket No. 0463.), filed March 22, 2012, filed Mar. 22, 2012. No. 13 / 554,483 “Hemodynamic Reser” by Grudic et al., Filed on Jul. 20, 2012, claiming priority to 07PR, hereinafter referred to as the “'426 application”). Continued as part of “ve Monitor and Hemodialysis Control” (Attorney Docket No. 0463.05, hereinafter referred to as “'483 Application”), all of which are incorporated herein by reference.

  [0005] The '483 application is also filed on March 4, 2010 by Grudic et al., US Provisional Patent Application No. 61 / 310,583, “Active Physical Perturbations to Enhance Intelligent Medical Monitoring” (Attorney Docket Number). US Patent Application No. 13 / 041,006 to Grudic et al., Filed Mar. 4, 2011, claiming priority over US Pat. No. 13 / 041,006, filed Mar. 4, 2011, claiming priority to US Pat. Continued as part of "Perturbations to Enhance Intelligent Medical Monitoring" (Attorney Docket No. 0463.04, hereinafter referred to as "'006 Application") Both are incorporated herein by reference. The '006 application was also filed on February 16, 2010 by Multon et al., US Provisional Patent Application No. 61 / 305,110, “Statistical, Noninspired Method for Predicting Intracranial Pressure” (Attorney Docket No. 0463.03PR). US Patent Application No. 13 / 028,140 ("'140 Application") by Grudic et al. Filed on Feb. 15, 2011, claiming priority to the "' 110 application"). Currently, U.S. Pat. No. 8,512,260) as a part of “Statistical, Noninvasive Measurement of Intracranial Pressure” (Attorney Docket No. 0463.03). It is, both incorporated herein by reference.

  [0006] The '140 application is filed with US Provisional Patent Application 61 / 252,978, filed October 19, 2009, and US Provisional Patent Application 61/166, filed April 3, 2009. , 499, U.S. Provisional Patent Application No. 61 / 166,486, and U.S. Provisional Patent Application No. 61 / 166,472, and U.S. Provisional Patent Application No. 61/109, filed Oct. 29, 2008. International Patent Application No. PCT / US2009 / 062119 ("'119 Application"), Long Term Active from Large Continuous Changing, filed October 26, 2009, claiming priority to 490. Part of Data Sets ”(agent reference number 04633.01 / PCT) It continues Te, are incorporated herein by reference either.

[0007] The disclosures of each of these commonly assigned applications / patents (collectively "related applications") are hereby incorporated by reference in their entirety for all purposes.
Statement on rights to inventions made under federal-supported research and development

[0008]
The present invention is license number 0535269 awarded by the National Science Foundation, grant number FA8650-07-C-7702 awarded by the Air Force Institute, and grant number W81XWH-09-C- awarded by the Army Medical Research Command. Made with government support under 1060 and W81XWH-09-1-0750. The government has certain rights in the invention.

Copyright statement
[0009]
Part of the disclosure of this patent document includes material that is subject to copyright protection. The copyright owner does not object to facsimile reproduction by anyone in the patent document or patent disclosure as described in the Patent and Trademark Office patent file or record, but otherwise reserves all copyrights. To do.

  [0010] The present disclosure relates generally to tools and techniques for medical monitoring, and more particularly to tools and tools that can provide rapid detection of bleeding before, during, and after infusion resuscitation. Regarding technology.

  [0011] Hemorrhagic shock induced by traumatic injury is a leading cause of death. The first hour after the injury is called the “golden time,” which can split life and death by recognizing and properly managing patients with significant continuous bleeding during that short period of time. Because there is. Significant bleeding is not always evident clinically. Many seriously injured patients have intracavitary bleeding, which means that bleeding from major organs or blood vessels is present in the chest or abdomen. There is no external evidence of bleeding, and as a result, doctors must look for suspicion of bleeding and clinical signs. In situations where imaging and laboratory tests are not generally available, changes in vital signs over time may be the only indication that a patient is bleeding. Thus, during "golden time", one must learn to recognize signs and symptoms of acute blood loss, then initiate resuscitation with fluids and frequently estimate patient fluid demands frequently.

  [0012] The problem is that humans cannot perceive subtle beat-to-beat vital signs changes that are indicative of bleeding. More importantly, humans are unable to detect the subtle changes in vital signs that are characteristic of hemodynamic decompensation or cardiovascular collapse, as predicted by hypotension with bradycardia.

  [0013] To further complicate the problem, humans have the innate ability to compensate for significant blood loss with little change in traditional vital signs. Therefore, it is difficult to detect blood loss using conventional vital sign monitoring techniques.

  [0014] Thus, there is a need for automated, non-invasive devices for early diagnosis, real-time monitoring, and tracking of blood loss, especially before, during, and after infusion resuscitation.

  [0015] A further understanding of the nature and advantages of particular embodiments may be gained by reference to the remaining portions of the specification and drawings, wherein the same reference numbers are used to refer to similar components. In some examples, the sublabel is associated with a reference number to indicate one of a plurality of similar components. When referring to a reference number without specifying an existing sublabel, it shall refer to all such similar components.

FIG. 2 is a schematic diagram illustrating a system for estimating compensatory reserves according to various embodiments. 1 is a schematic diagram illustrating a sensor system that can be worn on a patient's body, according to various embodiments. FIG. FIG. 5 is a process flow diagram illustrating a method for assessing blood loss according to various embodiments. FIG. 6 illustrates a technique for assessing blood loss according to various embodiments. FIG. 5 is a process flow diagram illustrating a method for estimating a patient's compensatory reserves and / or dehydration status according to various embodiments. FIG. 6 illustrates a technique for estimating and / or predicting a patient's compensatory reserve index, according to various embodiments. FIG. 5 is a process flow diagram illustrating a method for generating a model of a physiological state, according to various embodiments. FIG. 3 is a process flow diagram illustrating a method for performing rapid detection of bleeding before, during, and after infusion resuscitation according to various embodiments. 6 is an exemplary screen capture showing the display capability of a compensatory reserve monitor showing assessment of blood loss before, during, and / or after resuscitation by infusion according to various techniques. 6 is an exemplary screen capture showing the display capability of a compensatory reserve monitor showing assessment of blood loss before, during, and / or after resuscitation by infusion according to various techniques. 6 is an exemplary screen capture showing the display capability of a compensatory reserve monitor showing assessment of blood loss before, during, and / or after resuscitation by infusion according to various techniques. 6 is an exemplary screen capture showing the display capability of a compensatory reserve monitor showing assessment of blood loss before, during, and / or after resuscitation by infusion according to various techniques. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 6 is a graph illustrating rapid detection of bleeding before, during, and after resuscitation by infusion of a patient in a multi-organ traumatic clinical trial at Denver Health Medical Center, according to various embodiments. FIG. 2 is a generalized schematic diagram illustrating a computer system in accordance with various embodiments.

  [0027] Overview

  [0028] Various embodiments can detect bleeding before, during, and after resuscitation with an infusion. In one aspect, such detection can be performed non-invasively. In some embodiments, the detection is patient compensatory reserve index (“CRI”, also referred to herein as “cardiac reserve index” or “hemodynamic reserve index” (HDR)). Based on the calculation (or estimation) of. In other cases, the evaluation may be based on raw waveform data (eg, PPG waveform data) captured by a patented sensor (eg, a sensor described in a related application). In further cases, the combination of waveform data and calculated / estimated CRI can be used to calculate the effectiveness of resuscitation and / or the amount of infusion required for effective resuscitation.

  [0029] In other aspects, such functionality is in particular by systems and devices (such as cardiac reserve monitors), tools, techniques, methods, and software described in related applications, including the '483 application. Can be provided and / or integrated therewith. For example, the various operations described according to the methods disclosed by the related applications can be used in a method for assessing the effectiveness of resuscitation and / or a method for calculating the amount of infusion required for effective resuscitation. Similarly, such techniques can be performed by the system and / or implemented by software products described in related applications.

  [0030] One embodiment includes one or more sensors placed on a patient and one or more of the relevant parameters outlined above (in real time, after every heartbeat, or information is required. A system comprising a computer system (e.g., as described in a related application) that performs a method of using sensor data to estimate and predict (if any). Other embodiments comprise a computer system programmed to perform such a method, an apparatus including instructions for programming a computer to perform such a method, and / or such a method itself. Can do.

  [0031] Sensors include non-invasive blood pressure sensors such as Nexfin (BMEYE, BV) or Finometer (Finapress Medical Systems BV), invasive arterial blood pressure using arterial catheters, invasive central veins Pressure, invasive or non-invasive intracranial pressure monitor, EEG (electroencephalograph), heart monitor (EKG), transcranial Doppler sensor, transthoracic impedance plethysmography, pulse oximetry, sensor generating photoplethysmograph (PPG) waveform, Including but not limited to any of near infrared spectroscopy, electronic stethoscope, and / or the like.

  [0032] Although the '809 application describes several exemplary embodiments, the various embodiments are not limited to those described in the' 809 application. For example, FIG. 1 of the '809 application shows an exemplary sensor that can be used to collect waveform data for analysis, but other sensors can be used as well. Similarly, the '809 application describes several techniques for estimating the probability of blood loss. Many such techniques rely on an estimate of the patient's CRI that can be calculated using the techniques described in the '483 application. However, it should be understood that in various embodiments, other embodiments of estimating the probability of bleeding and / or estimating the CRI can be employed.

  [0033] Thus, in one aspect, the method uses techniques that include, but are not limited to, receiving data from such sensors and analyzing the data using the models described in the related applications. Such data can be analyzed. By way of example only, the model can be used to generate subject data for tests such as the LBNP test that can be used to predict or estimate CRI (or HDRI) values, as described in related applications, particularly the '483 application. Can be built using. From this calculated CRI value (or, in some embodiments, from the waveform data itself, alone or in combination with the CRI value), the patient can internalize before, during, and / or after resuscitation with an infusion. The probability of bleeding uses, for example, the technique described in the '809 application.

  [0034] For example, in one embodiment, the method captures waveform data from a patient using a sensor before, during, and / or after resuscitation with an infusion, and / or the patient's CRI value at these times Can be calculated. In some cases, changes in CRI values obtained during treatment can be used to estimate the probability that a patient is bleeding. For example, more fully described with respect to the clinical studies detailed in the '809 application, using standard deviations in CRI values during recording and / or differences in CRI values before, during, and / or after resuscitation by infusion As can be seen, the probability of bleeding can be estimated.

  [0035] Some embodiments further comprise normalizing the estimated probability of bleeding against scaling. For example, in some cases, an index from 0 to 1 can be used, where 0 indicates that the patient is not bleeding, 1 indicates that the patient is bleeding, and a value between 0 and 1 is , Based on an estimate calculated from the CRI value, indicates the relative probability that the patient is bleeding.

  [0036] In the following detailed description, several exemplary embodiments are described in further detail to enable those skilled in the art to practice such embodiments. The described examples are provided for the purpose of illustration and are not intended to limit the scope of the invention. In the present disclosure, it should be recognized that a node may be “virtual” or supported on a hypervisor or host system, or may be a physical node or network device in a network. In most cases, these figures show bridging a virtual path and possibly a node (virtual machine) between the path or two physical nodes. However, it should be understood that a “swap” of paths through an organization can occur in any combination, such as physical and / or virtual nodes, or physical and / or virtual links.

  [0037] In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the described embodiments. However, it will be apparent to those skilled in the art that other embodiments of the invention may be practiced without some of these specific details. In other instances, specific structures and devices are shown in block diagram form. Although several embodiments are described herein and various features are attributed to different embodiments, it is to be understood that the features described with respect to one embodiment can be incorporated into other embodiments. However, for similar reasons, other embodiments of the invention may omit such features, so that any single feature or features of any described embodiment may It should not be considered essential to the embodiment.

  [0038] Unless otherwise indicated, all numerical values used herein to express amounts and dimensions used, etc., are understood as being modified in all cases by the term "about". Should be. In this application, the use of the singular includes the plural unless specifically stated otherwise and the use of the terms “and” and “or” means “and / or” unless stated otherwise. Furthermore, the use of other forms such as “include” and “include” and “include” should be considered non-exclusive. Also, terms such as “element” or “component” encompass both elements and components comprising one unit and elements and components comprising two or more units, unless otherwise specified.

  [0039] Tools provided by various embodiments include, but are not limited to, methods, systems, and / or software products. Merely by way of example, a method may include one or more procedures, any or all of which are performed by a computer system. Correspondingly, embodiments can provide a computer system configured with instructions for performing one or more procedures in accordance with the methods provided by various other embodiments. Similarly, a computer program may comprise a set of instructions that can be executed by a computer system (and / or a processor therein) to perform such operations. In many cases, such software programs are encoded on physical, tangible, and / or non-transitory computer readable media (eg, optical and / or magnetic media, to name a few examples). Is done.

  [0040] In one aspect, a system can be provided that includes one or more sensors for obtaining physiological data from a patient and a computer system in communication with the one or more sensors. The computer system can comprise one or more processors and a non-transitory computer readable medium in communication with the one or more processors. The computer readable medium encodes a set of instructions executable by one or more processors to cause the computer system to receive physiological data from one or more sensors, analyze the physiological data, and cause the patient to bleed And at least one of an evaluation, a predicted value, or an estimated value indicating the probability that the patient is bleeding can be displayed on the display device.

  [0041] In another aspect, one or more sensors are used to monitor patient physiological data, analyze the physiological data with a computer system, and based on at least a portion of the analyzed physiological data A method can be provided comprising estimating a probability that a patient is bleeding. The method may further comprise displaying an indication on the display device of at least one of an estimate of the probability that the patient is bleeding, a predicted value, or an estimated value.

  [0042] In some cases, monitoring physiological data, analyzing physiological data, estimating the probability that a patient will bleed, or assessing, predicting, or estimating the probability that a patient will bleed One or more of displaying at least one of the instructions is performed in real time. In some cases, estimating the probability that a patient is bleeding may be based on one or more values of compensatory reserve index (CRI) values estimated based on received physiological data. Estimating the probability of bleeding can be provided. According to some embodiments, the one or more values of the CRI are physiological data that is at least one of received before, during, or after a resuscitation procedure with an infusion. Is estimated based on

  [0043] In some embodiments, the one or more values of the CRI may comprise multiple values of the CRI. In some cases, estimating the probability that a patient is bleeding may include an average value of CRI over a specific time period, at least some standard deviation of multiple CRI values, at least some distortion of multiple CRI values, multiple Based at least in part on one or more of a rate of change of at least some of the CRI values, a rate of rate of change of at least some of the plurality of CRI values, and / or a difference between at least some of the plurality of CRI values And estimating the probability that the patient is bleeding. In some cases, the indication is a value between 0 and 1. According to some embodiments, a value of 0 indicates that the patient is not bleeding and a value of 1 indicates that the patient is bleeding.

[0044] In some cases, estimating a patient's CRI comprises estimating a compensatory reserve index by comparing physiological data to a model constructed using the following equation:
Where CRI (t) is the compensatory reserve at time t, BLV (t) is the subject's intravascular volume loss at time t, and BLV _HDD is the subject's hemodynamic decompensation point Intravascular volume loss. In some embodiments, the physiological data comprises waveform data, and estimating the CRI determines that the one or more subjects are hemodynamic decompensated or near hemodynamic decompensation, or hemodynamic Comparing the waveform data with one or more sample waveforms generated by exposure to a series of conditions progressing towards decompensation, and monitoring the subject's physiological data.

  [0045] In some cases, the physiological data may comprise waveform data, and estimating the CRI may comprise comparing the waveform data to a plurality of sample waveforms, each of the sample waveforms being Corresponding to different values of the CRI to generate a similarity coefficient that represents the similarity between the waveform data and each of the sample waveforms, estimating the CRI further normalizes the similarity coefficient for each of the sample waveforms And summing the normalized similarity coefficients to generate an estimated CRI value for the patient.

  [0046] According to some embodiments, estimating the probability that a patient is bleeding is based on a fixed time history that monitors the patient's physiological data. Alternatively, estimating the probability that a patient is bleeding is based on a dynamic time history that monitors the patient's physiological data. In some examples, at least one of the one or more sensors includes a blood pressure sensor, an intracranial pressure monitor, a central venous pressure monitoring catheter, an arterial catheter, an electroencephalograph, a heart monitor, a transcranial Doppler sensor, a transthoracic impedance Selected from the group consisting of a plethysmograph, a pulse oximeter, a near-infrared spectrometer, a ventilator, an accelerometer, an electrical eye map, a transcutaneous glucometer, an electrolyte sensor, and an electronic stethoscope.

  [0047] As an example, in some embodiments, the physiological data may comprise at least one of blood pressure waveform data, plethysmograph waveform data, or photoplethysmograph (PPG) waveform data.

  [0048] In some cases, analyzing the physiological data can comprise analyzing the physiological data against an existing model. In some embodiments, the method can further comprise generating an existing model prior to analyzing the physiological data. In some cases, generating an existing model includes receiving data relating to one or more physiological parameters of the subject to obtain a plurality of physiological data sets, and using one of the subjects using a reference sensor. Alternatively, the method may comprise directly measuring a plurality of physiological conditions to obtain a plurality of physiological condition measurements and correlating the received data with the subject's physiological condition measurements. According to some embodiments, the one or more physiological conditions comprises a decrease in circulatory system volume.

  [0049] In some examples, the method may further include inducing a reduced circulatory volume physiological condition in the subject. In some cases, inducing the physiological condition comprises at least one of exposing the subject to lower body negative pressure (LBNP) and / or exposing the subject to dehydration. In some embodiments, the one or more physiological conditions include at least one of a cardiovascular collapse or near-cardiovascular collapse condition, a normal fluid volume condition, a circulatory blood volume condition, and / or a dehydration condition. Can be provided.

[0050] According to some embodiments, correlating the received data with the physiological condition measurements of subjects, the signal s _1, s 2 for each of the one or more results o _k,. . . , To identify the most predictive set of signals S _k from the set of s _D, probabilistic predictive model
Autonomously learning a set of signals and signals s ₁ , s ₂ ,. . . , S _D example values and corresponding results o ₁ , o ₂ ,. . . , O _K can be provided to repeat the autonomous learning operation step by step. Here, the most predictive set of signals S _k, corresponds to the first data set representing a first physiological parameter, each of the one or more results o _k, represents a physiological condition measurements ,
Is the predicted value of the result o _k resulting from the model M _k to be used as an input value obtained from a set of signal S _k.

  [0051] In yet another aspect, an apparatus can be provided that comprises a non-transitory computer readable medium encoded with a set of instructions executable by one or more computers, the set of instructions comprising: Receive physiological data from one or more sensors, analyze the physiological data, estimate the probability that the patient is bleeding, and evaluate, predict, or estimate the probability that the patient is bleeding At least one is displayed on the display device.

  [0052] Various modifications and additions may be made to the described embodiments without departing from the scope of the present invention. For example, although the embodiments described above relate to particular features, the scope of the invention includes embodiments having different combinations of features and embodiments that do not include all of the features described above.

  [0053] Compensatory reserve index (CRI)

  [0054] Various embodiments can assess the effectiveness of ingestion hydration, which, without limitation, leads to better hydration or optimal hydration Can be defined to maintain In one aspect, optimal hydration is probably referred to as the patient's Compensatory Reserve Index (“CRI”, herein the “cardiac reserve index” or “hemodynamic reserve index” (HDR), These should all be considered synonymous for purposes of this disclosure) and can be defined as an infusion state that maximizes the performance index / measurement indicated. (For convenience, the term “patient” is used herein, but this description should not be considered limiting, as various embodiments may be used prior to athletics or training. This is because it can be used both in the clinical field and outside the clinical field by middle or later athletes, people in daily activities, soldiers on the battlefield, etc. Therefore, the term “patient” used in the present specification. The term should be interpreted broadly and should be considered synonymous with "person"). In other cases, the assessment may be based on raw waveform data (eg, PPG waveform data) captured by a patented sensor (eg, a sensor described below and in related applications). In further cases, the combination of waveform data and calculated / estimated CRI can be used to calculate the effectiveness of hydration and / or the amount of infusion required for effective hydration. In other aspects, such functionality is provided by systems, devices (such as cardiac reserve monitors and / or wrist-worn sensor devices), tools, techniques, methods, and software described below and in related applications. Can be and / or integrated with them.

  [0055] For example, one set of embodiments provides a method. Exemplary methods may comprise monitoring patient physiological data using one or more sensors. The method can further comprise analyzing the physiological data using a computer system. Many different types of physiological data are monitored by various embodiments, including but not limited to blood pressure waveform data, plethysmograph waveform data, photoplethysmograph (PPG) waveform data (such as those generated by a pulse oximeter). And / or can be analyzed. In one aspect of some embodiments, analyzing the physiological data can comprise analyzing the data against an existing model. In some cases, the method can further comprise assessing the effectiveness of the hydration effort and / or displaying the assessment of the effectiveness of the hydration effort (eg, on a display device). Such assessments include, but are not limited to, estimation of efficacy at this time, prediction of efficacy at some future time, estimation and / or prediction of the volume of infusion required for effective hydration, patient May include an estimate of the probability that the fluid requires an infusion.

  [0056] An apparatus according to yet another set of embodiments comprises a computer-readable medium having encoded thereon a set of instructions executable by one or more computers to perform one or more operations. Can do. In some embodiments, the set of instructions may comprise instructions for performing some or all of the operations of the methods provided by the particular embodiment.

  [0057] A system according to yet another set of embodiments may comprise one or more processors and a computer-readable medium in communication with the one or more processors. A computer-readable medium may encode a set of instructions that can be executed by a computer system to perform one or more operations, such as the set of instructions described above, to name one example. In some embodiments, the system can further comprise one or more sensors and / or treatment devices, either or both of which are in communication with and / or controlled by the processor. There is a possibility. Such sensors include, but are not limited to, blood pressure sensors, intracranial pressure monitors, central venous pressure monitoring catheters, arterial catheters, electroencephalographs, heart monitors, transcranial Doppler sensors, transthoracic impedance plethysmographs, pulses It can include oximeters, near infrared spectrometers, ventilators, accelerometers, electrograms, transcutaneous glucometers, electrolyte sensors, and / or electronic stethoscopes.

  [0058] CRI to assess blood loss before, during, and / or after resuscitation by infusion

  [0059] A set of embodiments is a method that can be used (eg, before, during, and / or after resuscitation by infusion) to quickly and accurately assess patient blood loss, often non-invasively, Provide systems and software. Such assessments include, but are not limited to, estimation of efficacy at this time, prediction of efficacy at some future time, estimation and / or prediction of the volume of infusion required for effective hydration, patient Estimation of the probability of needing infusion, estimation and / or prediction of blood loss (eg, before, during, and / or after resuscitation by infusion), and the like. In a particular set of embodiments, an apparatus that can be worn on the patient's body can include one or more sensors that monitor the patient's physiological parameters. The device (or a computer in communication with the device) analyzes the data captured by the sensor and compares that data with a model (which can be generated according to other embodiments) and described in more detail in the '426 application. As is done, the effectiveness of hydration can be assessed and / or blood loss (eg, before, during, and / or after resuscitation by infusion) can be assessed.

  [0060] Various embodiments can measure a number of different physiological parameters from a patient, and analysis of these parameters determines which parameters are measured (and which, depending on the model generated) Is the most predictive of the effectiveness of hydration, including the probability of need for hydration and / or the amount of infusion needed, or the most predictive of blood loss). In some cases, the parameters themselves (eg, continuous waveform data captured by a photoplethysmograph) are analyzed against the model to validate hydration (eg, before, during, and / or after resuscitation by infusion). Sex assessment or blood loss assessment can be performed. In other cases, physiological parameters can be obtained from the captured data and these parameters can be used. By way of example only, estimating the value of CRI using direct physiological data (captured by a sensor) as described further below and in the '483 application (already incorporated by reference) This value of CRI can be used to assess the effectiveness of hydration (before, during, and / or after resuscitation by infusion) and / or to assess blood loss. In still other cases, the derived CRI value and raw sensor data can be used together to make such an evaluation.

  [0061] For example, the '483 application describes a compensatory reserve monitor (also called a cardiac reserve monitor or a hemodynamic reserve monitor) that can estimate a patient's compensated reserve. In one aspect, the monitor can quickly, accurately, and / or determine in real time the probability that a patient is bleeding. In another aspect, the device can simultaneously monitor a patient's compensatory reserve by tracking the patient's CRI to properly and effectively guide hydration and ongoing patient care. The same device (or similar device) evaluates the effectiveness of hydration based on the monitored CRI value and / or (eg, by infusion), as described in further detail in the '426 application. It can include advanced features that quickly assess blood loss (before, during, and / or after resuscitation).

  [0062] CRI is a hemodynamic parameter that indicates the individual ratio of intravascular fluid reserve that remains prior to the onset of hemodynamic decompensation. The CRI has a value in the range of 1 to 0, a value close to 1 is associated with normal blood pressure (normal circulation volume), and a value close to 0 is an individual specific circulation volume at which hemodynamic decompensation occurs. Related.

[0063] The formula for CRI at a certain time "t" is given by:

(Formula 1)

[0064] BLV (t) is the human intravascular volume loss at time “t” (“BLV”, also referred to as “blood loss volume” in the related application), and BLV _HDD is a hemodynamic decompensation (HDD) When entering, it is a human intravascular volume loss. Hemodynamic decompensation is generally defined as occurring when systolic blood pressure falls below 70 mmHg. This level of intravascular volume loss is individually specific and varies from subject to subject.

[0065] The lower body negative pressure (LBNP) of the linear or non-linear relationship λ with intravascular volume loss is as follows:

BLV = λ · LBNP (Formula 2)

[0066] Can be used to estimate the CRI of an individual undergoing an LBNP experiment as follows:
(Formula 3)

[0067] LBNP (t) is the LBNP level that the individual is experiencing at time “t”, and the LBNP _HDD is the LNPB level at which the individual enters hemodynamic decompensation.

[0068] Either CRI data, raw (or processed) sensor data, or both can be used to assess the effectiveness of hydration according to various embodiments. In one embodiment, the assessment of blood loss (BL) can be expressed as a value between 0 and 1, with BL = 1 ensuring blood loss, BL = 0 with no blood loss, BL If is a value between 1 and 0, the value indicates the probability of blood loss probably due to blood loss before, during and / or after resuscitation by infusion. (Of course, other embodiments can scale the value of BL differently). In one aspect of some embodiments, the estimate is generally represented by the following formula:

BL = f _BL (CRI _t , FV _t , S _t ) (Formula 4)

[0069] Where BL is a measure or estimate of bleeding, f _BL (CRI _t , FV _t , S _t ) is described by an empirically generated model, eg, with respect to FIG. 4 below. CRI _t is a CRI value (can range from a single CRI value to a CRI value of several hours), an algorithm embodied using techniques and / or techniques described in related applications. a time history, FV _t is the infusion of time history given to the patient (which may be in the range of several hours values from a single value), S _t is described elsewhere herein As is a time history of raw sensor values, such as physiological data measured by the sensor (can range from a single value to a few hours value).

[0070] Function of Formula 4 were collected from human subjects at various levels _{BL (CRI t, FV t,} S t) time history data, _(CRI t for the current patient being monitored, FV It is similar to the format of the CRI model in the sense that it is compared with the time history of _{t 1} , S _t ), but is not limited to this. In that case, the estimated BL of the current patient is the closest (CRI _t , FV _t , S _t ) space with the previously collected data.

[0071] Although Equation 4 is a general formula for BL, various embodiments may use a subset of the parameters considered in Equation 4. For example, in one embodiment, the model may only consider the amount of infusion and CRI data without considering raw sensor input. In that case, BL can be calculated as follows.

BL = f _BL (CRI _t , FV _t ) (Formula 5)

[0072] Similarly, some models may estimate BL based on sensor data rather than initially estimating CRI, where BL may be expressed as:

BL = f _BL (FV _t , S _t ) (Formula 6)

  [0073] The choice of parameters to use in modeling the BL is arbitrary and depends on which parameters are shown (eg, using the technique of FIG. 4 below) to yield the best prediction of the BL can do.

[0074] In another aspect, the effectiveness of hydration can be assessed by estimating or predicting the volume V of infusion required for effective hydration of the patient. This amount V indicates the infusion volume required for complete hydration if treatment has not yet begun and / or the remaining amount for fully effective hydration if treatment is ongoing. be able to. Similar to BL, the value of V can be estimated / predicted using the modeling techniques described herein and in related applications. In the general case, V can be expressed as:

V = f _V (CRI _t , FV _t , S _t ) (Expression 7)

Where V is the estimated amount of infusion the patient needs to prevent excessive or underhydration and f _V (CRI _t , FV _t , S _t ) is, for example, with respect to FIG. And / or an algorithm embodied by an empirically generated model using techniques described in related applications, where CRI _t is the time history of CRI values and FV _t is given to the patient a volume of infusion of time history, S _t is the time history of the physiological data received from one or more sensors.

[0075] Similar to the estimate of BL, in various embodiments, a subset of the parameters used in the general equation of Equation 7 may be used. Accordingly, various embodiments can calculate V as follows.

V = f _V (CRI _t , FV _t ) (Formula 8)
Or V = f _V (FV _t , S _t ) (Formula 9)

[0076] Yet another way of assessing the effectiveness of hydration (which may include assessing the need for hydration) is to estimate the probability P _f that the patient needs an infusion, This probability can be used to estimate the likelihood that the patient will need hydration if treatment has not begun and / or if hydration treatment is ongoing, further hydration may be necessary Can be estimated. This probability value can be expressed, for example, as a percentage, as a decimal value between 0 and 1, and can be estimated using the following equation:
(Formula 10)
Where P _f is the estimated probability that the patient needs an infusion,
Is a relationship derived on the basis of empirical research, CRI _t is the time history of the CRI value and the S _t is the time history of the physiological data received from one or more sensors. Again, this general representation can be used in various embodiments using a subset of the parameters in the general representation as follows.
(Formula 11)
Or
(Formula 12)

[0077] In any estimate of BL, V, or P _f , the function f represents a relationship derived based on empirical studies. In a set of embodiments, for example, subjecting various sensor data to subjects under hydration efforts, under and / or after bleeding, during bleeding, or other conditions that may simulate such a situation. Can be collected from. This sensor data can be used to estimate various assessments of hydration effects using, for example, the methods described below with respect to FIGS. 2 and 3, a technique similar to that of FIG. 4 below. Can be used to analyze the model for development.

[0078] CRI, BL, V, and / or P _f measurements may be useful in a variety of clinical settings including, but not limited to: 1) acute blood loss due to injury or surgery, 2) blood Acute circulation loss due to dialysis (also called dialysis hypotension), 3) Acute circulation loss due to various causes of dehydration (eg, reduced intake, vomiting, dehydration, etc.). Changes in CRI may foresee other conditions including, but not limited to, blood pressure, general fatigue, overheating, and certain disease changes. Thus, tools and techniques for estimating and / or predicting CRI can have a variety of uses in clinical settings, including but not limited to diagnosing such conditions.

[0079] In addition, CRI, BL, V, and / or P _f measurements may be applicable outside the clinical setting. For example, athletes should monitor (eg, using a wrist-worn hydration monitor) to ensure optimal performance (and overall health and recovery) before, during, or after competition or training Can do. In other situations, those who are concerned about overall health can use a similar hydration monitor to ensure that sufficient (but not too much) infusions are obtained and diseases (eg, To ensure that dehydration does not occur due to vomiting, diarrhea), the sick infant or adult can be monitored during the illness. Similarly, soldiers in the field (especially harsh conditions) can be monitored for optimal operational responsiveness.

  [0080] In various embodiments, a hydration monitor, a compensatory reserve monitor, a wrist-mounted sensor device, and / or another integrated system, as described in further detail herein and in related applications, include: However, the present invention is not limited to these functions.

  [0081] A. Estimating and / or displaying intravascular volume loss for hemodynamic decompensation (or cardiovascular collapse).

  [0082] B. Estimating, predicting, and / or displaying a patient's compensatory reserve as an index that is proportional to an approximate measure of intravascular volume loss for CV collapse, recognizing that each patient has its own spare capacity.

  [0083] C.I. Reference value at normal body fluid volume (eg, CRI = 1), representing a state in which the patient has normal blood flow, minimum value (eg, CRI = 1), lack of circulatory reserve, and patient experiencing CV collapse , CRI = 0), and / or estimating, predicting and / or predicting a patient's compensatory reserve indexed by a normative value in an overvalue (eg, CRI> 1) representing a condition in which the patient has excessive circulation The patient's normalized compensatory reserve can be displayed in a continuum between the minimum and maximum values (possibly different symbols depending on where the patient is located on the continuum) And / or labeled by color).

  [0084] D. To determine and / or display the probability that bleeding or intravascular volume loss has occurred.

  [0085] E.E. To display an index of the occurrence and / or progression of intravascular volume loss, as well as other measures of reserve, such as trend lines.

  [0086] F. Estimating the patient's current blood pressure and / or predicting the patient's future blood pressure.

  [0087] G.G. To assess the current effects of resuscitation efforts through infusions.

  [0088] H.C. Predict the future effects of resuscitation efforts with infusions.

  [0089] I.D. Estimating and / or predicting the volume of infusion necessary for effective resuscitation.

  [0090] J. et al. Estimate the probability that a patient will need an infusion.

  [0091] K.K. To estimate the hydration status of a patient or user.

  [0092] L. Predict the future hydration status of the patient or user.

  [0093] M.M. Estimate and / or predict fluid intake required for proper hydration of the patient or user.

  [0094] N. Estimate the probability that a patient will dehydrate.

[0095] In various embodiments, the CRI, BL, V, and / or _Pf is based on (i) a fixed time history of patient monitoring (eg, 30 seconds or 30 heartbeat windows), (ii) patient monitoring Dynamic time history (eg, 200 minute monitoring, the system uses all sensor information collected during that time to improve and improve CRI estimates, hydration effectiveness assessments, etc. (Iii) based on (iii) establishing a baseline estimate if the patient has normal blood flow (no volume loss has occurred) and / or (iv) if the patient has normal blood It can be estimated based on the NO baseline estimate for the flow rate.

  [0096] Certain embodiments may also be based on analysis of patient status (including estimated / predicted blood pressure, bleeding probability, dehydration status, and / or patient estimate and / or predicted CRI). Can also be recommended. Treatment options include, but are not limited to, hemodynamics, ventilator adjustment, IV infusion adjustment (eg, control of IV pump flow rate or IV infusion rate), blood or blood product infusion, drug changes , Patient position changes, surgical treatments, and the like.

  [0097] As an example, certain embodiments can be used to control an IV drip, IV pump, or rapid infuser. For example, embodiments may estimate the probability that a patient will require an infusion and activate such a device in response to the estimate (or attach a such device to the patient and activate the device to the clinician) Can be instructed). The system can then monitor the progress of the hydration effort (through continuous or periodic assessment of hydration effectiveness) and increase or decrease infusion or flow accordingly.

  [0098] As another example, certain embodiments can be used as input for a hemodialysis procedure. For example, certain embodiments can predict how much intravascular (blood) volume can be safely removed from a patient during a hemodialysis process. For example, embodiments may provide instructions to a human operator of a hemodialysis machine based on an estimated or predicted value of a patient's CRI. Additionally and / or alternatively, such embodiments can be used to continuously self-adjust the ultrafiltration rate of a hemodialyzer, thereby dialysis hypotension and associated morbidity. Avoid rates entirely.

  [0099] As yet another example, certain embodiments are used to estimate and / or predict dehydration status (and / or amount of dehydration) in an individual (eg, trauma patient, athlete, home elderly, etc.). And / or can provide treatment (by recommending to the therapist or by directly controlling the appropriate treatment device). For example, an analytical model can be used to determine the relationship between CRI (and / or any other physiological phenomenon that can be measured and / or estimated using the techniques described herein and related applications) and dehydration. Where indicated, embodiments can apply the model to estimate the dehydration status of the patient using the techniques described herein.

  [0100] Certain exemplary embodiments

  [0101] Reference will now be made to the embodiments illustrated in the drawings. FIGS. 1-10 illustrate some of the features of methods, systems, and devices for rapid detection of bleeding before, during, and after infusion resuscitation as described above. FIGS. 1-8 illustrate some (but not limited to) certain exemplary features of methods, systems, and devices for rapid detection of bleeding before, during, and after resuscitation with an infusion. 9A through 9H show the implementation of rapid detection of bleeding before, during, and after resuscitation by patient infusion in clinical trials, and FIG. System and hardware implementation is shown. The methods, systems, and apparatuses illustrated by FIGS. 1-10 illustrate examples of different embodiments that include various components and steps, which can be considered as alternatives, or in each other in various embodiments. Can be used with. The description of the illustrated methods, systems, and apparatus shown in FIGS. 1-10 is provided for purposes of illustration and should not be considered as limiting the scope of the different embodiments.

  [0102] Referring to the drawings, FIG. 1A provides an overview of the system provided by certain embodiments. The system includes a computer system or computing device 100 that communicates with one or more sensors 105 configured to obtain physiological data from a subject (eg, an animal or human subject or patient) 110. In one embodiment, the computer system 100 comprises a Lenovo THINKPAD X200, RAM4GB with a Microsoft WINDOWS® 7 operating system and is programmed with software that performs the computational methods outlined herein. The calculation method can be implemented in MATLAB 2009b and the C ++ programming language. A more general example of a computer system 100 that can be used in some embodiments is described in further detail below. More generally, however, the computer system 100 can be any system of one or more computers capable of performing the techniques described herein. In certain embodiments, for example, the computer system 100 reads values from the physiological sensors 105, generates models of physiological conditions from the sensors, and / or uses such models to provide individual-specific Generate estimates, predictions, or other diagnostic values, display the results, recommend and / or perform therapeutic treatment as a result of the analysis, and / or archive (learn) these results Can be used for future model building and forecasting.

  [0103] The sensor 105 can be any of a variety of sensors (including but not limited to those described herein) for obtaining physiological data from a subject. Exemplary sensor suites include a Finometer sensor for obtaining a non-invasive continuous blood pressure waveform, a pulse oximeter sensor, and an analog-to-digital board for connecting a sensor (pulse oximeter and / or finometer) to the computer system 100. (National Instruments USB-9215A 16 bits, 4 channels). More generally, in one embodiment, the one or more sensors 105 use continuous physiological waveform data, such as continuous blood pressure, using, for example, one or more of the techniques described herein. Can be obtained. The input from sensor 105 can constitute a continuous data signal and / or result that can be generated and / or applied to the prediction model as described below.

  [0104] In some cases, the structure or system may be controlled by the computer system 100 to deliver a therapeutic procedure according to recommendations developed by analysis of patient physiological data ("herein" Also referred to as a “physiological assist device”). In certain embodiments, the treatment device comprises a hemodialysis device (also referred to as a hemodialysis machine) that can be controlled by the computer system 100 based on the patient's estimated CRI, as described in more detail below. Can do. Additional examples of treatment devices in other embodiments include cardiac assist devices, ventilators, automatic implantable cardioverter defibrillators (AICDs), pacemakers, extracorporeal membrane oxygenation circuits, positive airway pressure (PAP) devices (limited) (But not continuously positive airway pressure (cPAP) devices), anesthesia devices, integrated lifesaving systems, medical robots, infusions and / or therapeutic compounds can be provided intravenously (eg, by intravenous injection) and Intraarterial pumps, intravenous infusions, rapid injectors, and / or heating / cooling blankets, etc. can be included.

  [0105] FIG. 1B shows in greater detail an exemplary sensor device 105 that may be used in the system 100 described above. (Of course, the illustrated sensor device 105 of FIG. 1B is not intended to be limiting, and various embodiments include sensors described elsewhere in this disclosure and related applications. Any sensor that captures appropriate data without limitation can be used). The illustrated sensor device 105 is designed to be worn on a patient's wrist, and thus blood pressure during (eg, but not limited to, athletics or training, daily activities, military training or actions, etc. Any person that may be beneficial to monitor for a variety of reasons, including assessing hydration and / or) can be used both clinically and locally. In one aspect, the sensor device 105 can function as an integrated hydration monitor, which evaluates hydration, displays instructions for evaluation, and exercises as described herein. Therapeutic behavior can be recommended based on, for example, an evaluation of form factors that can be worn during daily activities.

  [0106] Accordingly, the exemplary sensor device 105 (hydration monitor) includes a finger cuff 125 and a wrist unit 130. The finger cuff 125 includes a fingertip sensor 135 (in this case, a PPG sensor) that captures data based on the patient's physiological condition, such as PPG waveform data. The sensor 135 communicates with the I / O unit 140 of the wrist unit 130 and supplies the output from the sensor 135 to the processing unit 145 of the wrist unit 130. Such communication may be wired (eg, via a standard or dedicated connector such as USB on the wrist unit 130) and / or wireless (eg, Bluetooth®, such as Bluetooth® Low Energy (BTLE)). ), Near Field Communication (NFC), WiFi, or any other suitable wireless technology).

  [0107] In various embodiments, the processing unit 145 may have different types of functions. For example, in some cases, the processing unit 145 stores the data prior to transmitting the data via the I / O unit 140 to the monitoring computer 100 that can perform data analysis to control the treatment device 115 or the like. And / or can simply operate to organize. However, in other cases, the processing unit 145 (eg, some or all of the components described below in connection with FIG. 10 and / or functions attributable to the computer 100 of FIGS. 1A and 1B). The processing unit 145 performs on-board data analysis, for example, to estimate and / or predict the patient's current and future blood pressure. can do. Thus, wrist unit 105 includes, but is not limited to, estimates and / or predictions (eg, CRI, blood pressure, hydration status, etc.) and / or data (heart rate, pulse rate) captured by the sensor. Display), which can display any output described herein, including

  [0108] In some cases, wrist unit 130 can include a wrist strap 155 that allows the unit to be worn on the wrist, similar to a wristwatch. Of course, other options for facilitating transport of the patented sensor device 105 are available. More generally, the sensor device 105 may not include all of the components described above and / or may combine and / or reconfigure various components, and the embodiment shown in FIG. It should be considered illustrative in nature and not limiting.

  [0109] FIGS. 2A, 2B, 3A, 3B, 4, and 5 illustrate methods and screen displays according to various embodiments. The methods of FIGS. 2A, 2B, 3A, 3B, 4 and 5 are shown as different methods for ease of explanation, but the various techniques and procedures of these methods are as follows: Can be combined in any suitable manner, and in some embodiments, the methods shown in FIGS. 2A, 2B, 3A, 3B, 4 and 5 are interoperable and / or single Can be considered as part of the method. Similarly, the techniques and procedures are shown and / or described in a specific order for illustration, and the specific procedures may be reordered and / or omitted within the scope of the various embodiments. Please understand that. In addition, the method illustrated in FIGS. 2A, 2B, 3A, 3B, 4, and 5 is performed by the computer system 100 of FIG. 1 (other components of the system, such as the sensor 105 of FIGS. 1A and 1B). (And, in some cases, described below with respect to them), these methods may be implemented using any suitable hardware implementation. Similarly, the computer system 100 of FIG. 1 (and / or other components of such a system) may follow the methods shown in FIGS. 2A, 2B, 3A, 3B, 4, and 5 (eg, Although it can operate (by executing instructions implemented on a computer-readable medium), the system 100 can also operate according to other modes of operation and / or perform other suitable procedures.

  [0110] By way of example only, a method may include one or more procedures, any or all of which are performed by a computer system. Correspondingly, embodiments can provide a computer system configured with instructions for performing one or more procedures in accordance with the methods provided by various other embodiments. Similarly, a computer program may comprise a set of instructions that can be executed by a computer system (and / or a processor therein) to perform such operations. In many cases, such software programs are encoded on physical, tangible, and / or non-transitory computer readable media (eg, optical and / or magnetic media, to name a few examples). Is done.

  [0111] As a non-limiting example, various embodiments can comprise a method of using sensor data to assess patient blood loss. FIG. 2 illustrates an exemplary method 200 according to various embodiments. The method 200 can comprise generating a model in a computer system that can analyze patient data, for example, to estimate and / or predict various physiological conditions (block 205). In a general sense, generating the model can comprise receiving data regarding a plurality of more physiological parameters of the subject to obtain a plurality of physiological data sets. Such data includes, but is not limited to, PPG waveform data and / or other types of sensor data showing examples, including but not limited to data captured by other sensors described herein and related applications. be able to.

  [0112] Generating the model can further comprise directly measuring one or more physiological states of the subject using a reference sensor to obtain a plurality of physiological state measurements. The one or more physiological conditions can include, but are not limited to, various volume states of blood loss and / or resuscitation by infusion and / or various states of hydration and / or dehydration. (In other embodiments, the different conditions can include a condition of excessive blood volume, a normal fluid volume condition, and / or a condition of cardiovascular collapse (or a condition close to cardiovascular collapse) and / or For example, it may include a simulated state through the use of an LBNP device). Other physiological conditions that can be used to generate the model are described elsewhere in this specification and related applications.

  [0113] Generating the model can further comprise correlating the physiological state with the measured physiological parameter. There are various techniques for using these general functions to generate models according to different embodiments. One exemplary technique for generating a model of a general physiological state is described below with reference to FIG. 4, which uses machine learning algorithms to measure measured physiological parameters. Various blood volume conditions, including hydration (such as PPG waveform data, as an example) and physical conditions (including known amounts of blood loss and / or resuscitation with known amounts of infusion) And / or techniques for optimizing the correlation between the various states of dehydration and the like. However, it should be understood that any suitable technique or model may be used in accordance with various embodiments.

  [0114] Several physiological states can be modeled and a number of different conditions can be imposed on the subject as part of the model generation. For example, physiological conditions that can be induced in a subject (or monitored spontaneously) include, but are not limited to, circulatory system volume reduction, a known amount of blood loss, a specific amount of infusion added to the blood volume, dehydration, Including cardiovascular collapse or a state close to cardiovascular collapse, normal body fluid volume, excessive circulating blood volume, hypotension, hypertension, and / or normal blood pressure.

  [0115] By way of example only, in one set of embodiments, various physiological parameters of multiple subjects can be measured. In some cases, subjects may receive various levels of blood loss (actual or simulated) or intravenous infusion. Using the method described below with respect to FIG. 4 (or other similar techniques, many of which are described in related applications), the system can determine which sensor information is between subjects at different blood loss / addition levels. It can be determined whether to most effectively distinguish.

  [0116] In addition to and / or instead of using direct (eg, raw) sensor data to build such a model, some embodiments provide data derived from sensor data. A model can be built based on By way of example only, one such model may use a subject's CRI value at different blood loss and / or volume addition conditions as an input value. Thus, the process of generating a model may comprise first building a CRI model and then building a hydration effectiveness model from that model. (In other cases, the hybrid model may consider both raw sensor data and CRI data).

  [0117] The CRI model can be generated in various ways. For example, in some cases, one or more subjects may receive LBNP. In an exemplary case, LBNP data is collected from subjects exposed to progressively low levels of LBNP until hemodynamic decompensation, at which time LBNP is found and the subject recovers. Each level of LBNP represents an additional amount of blood loss. During these tests, physiological data (including but not limited to waveform data such as continuous non-invasive blood pressure data) may be collected before, during and / or after application of LBNP. it can. As described above, the relationship between LBNP and intravascular volume loss (represented by Equation 2) can be identified, and this relationship can be used to estimate CRI. Thus, LBNP studies form a framework for the development of hemodynamic parameters, referred to herein as CRI, and can be used to generate models for this parameter.

  [0118] More generally, to generate such a model, a reduced physiological volume of the circulatory system, such as at a point near cardiovascular collapse (hemodynamic decompensation) or cardiovascular collapse. Several different techniques for inducing states can be used. As mentioned above, this condition can be triggered using LBNP. In some cases, dehydration can be used to induce this condition, as in the studies described below. Other techniques are possible. Similarly, in various embodiments, CRI models can be generated using data collected from subjects with normal fluid volume status, dehydration, circulating blood volume, and / or other conditions.

  [0119] At block 210, the method 200 may comprise monitoring the patient's physiological data using one or more sensors. As described above, various physical parameters can be monitored invasively and / or noninvasively, depending on the nature of the patient's anticipated physiological condition. In one aspect, monitoring one or more physical parameters can comprise receiving continuous waveform data that can be sampled as needed, eg, from a physiological sensor. Such data may include, but is not limited to, plethysmographic waveform data, and / or PPG waveform data (such as that generated by a pulse oximeter).

  [0120] The method 200 may further comprise analyzing the physiological data with a computer system (eg, the monitoring computer 100 and / or processing unit 135 of the sensor unit described above) (block 215). In some cases, the physiological data is against an existing model (which may be generated as described above and can be updated based on the analysis as described in more detail below and in related applications). Analyzed.

[0121] As an example only, in some cases, sensor data can be directly analyzed against the generated model to assess the effectiveness of hydration (as described above, estimating the current value). And / or predicting future values for any or all of BL, V, and / or P _f ). For example, sensor data can be compared to determine similarity to a model that estimates and / or predicts any of these values. By way of example only, an input waveform captured by a sensor from a patient can be compared to a sample waveform generated by a model of each of these values.

[0122] For example, the technique 200 'provides one method for deriving an estimate of BL, according to some embodiments. Technology 200 ′ is presented as an example and this technology 200 ′ estimates BL from raw sensor data, but using similar techniques, from raw sensor data, CRI data, and / or a combination of both. Note that BL, V, and / or P _f can be estimated or predicted. For example, one model generates a first estimate of BL from the raw sensor data, a second estimate of BL from the estimated CRI value, and these to generate a hybrid BL estimate Estimates can be combined (in either weighted or unweighted manner).

  [0123] The illustrated technique 200 ′ includes arterial waveform data such as, but not limited to, waveform data (eg, continuous PPG waveform and / or continuous non-invasive blood pressure waveform) for a specified period, such as 32 heart rates. Sampling (any of the data described herein and related applications) that is not (block 270). The sample is compared to a plurality of waveforms of reference data corresponding to BL values (block 275), in this case in the range of 0 to 1 using the scale described above (but using any suitable scale) To do). These reference waveforms are derived as part of a model developed using the algorithms described in this specification and related applications and can be the result of experimental data and the like. In essence, these reference waveforms reflect the relationship f from Equation 6.

  [0124] According to technique 200 ', the samples are compared with waveforms corresponding to BL = 1 (block 275a), BL = 0.5 (block 275b), and BL = 0 (block 275c) as shown. can do. (As indicated by the circles in FIG. 2B, any number of sample waveforms can be used for comparison, eg, if there is a non-linear relationship between measured sensor data and BL values, the sample waveform Can be a better comparison). From the comparison, a similarity factor is calculated (eg, using least squares or similar analysis) to represent the similarity between the sampled waveform and each of the reference waveforms (block 280). These similarity factors can be normalized (if appropriate) (block 285) and summed (block 290) to produce an estimated BL value for the patient (block 295). ).

  [0125] In other cases, similar techniques can be used to analyze data for a model based on parameters derived directly from sensor measurements. (In one aspect, such an operation may be derived by, for example, CRI or the like by analyzing sensor data for a first model and then analyzing parameters derived for a second model. Can be essentially iterative by generating parameters).

  [0126] For example, FIG. 3A assesses (in some embodiments) the effectiveness of hydration (including the probability that an infusion is needed and / or the estimated amount of infusion needed for effective hydration). To calculate the patient's CRI that can be used as a parameter that can be analyzed to assess and / or blood loss (eg, before, during, and / or after resuscitation by infusion) A method 300 is shown. The method 300 uses, for example, techniques such as those described above and in the '483 application to generate a CRI model (block 305), monitor physiological parameters (310), and monitored physics. Analyzing the global parameters (block 315).

  [0127] Based on this analysis, the method 300, in an exemplary embodiment, includes estimating the compensatory reserve of the patient at the computer system based on the analysis of the physiological data (block 320). In some cases, the method may further comprise predicting a compensatory reserve of the patient at one or more future time points with a computer system based on the analysis of the physiological data (block 325). Operations that predict values for future parameters can be similar to operations that estimate current values, but in a predictive situation, the model applied applies diagnostic data to the subject rather than simultaneous values. May correlate with subsequent values of the parameter. Of course, it is worth noting that in some embodiments, the same model can be used to estimate current values and predict future values of physiological parameters.

  [0128] The patient's estimated and / or predicted compensatory reserve can be based on a number of factors. Merely by way of example, in some cases, the estimated / predicted compensatory reserve can be based on a fixed time history that monitors patient physiological data and / or a dynamic time history that monitors patient physiological data. . In other cases, the estimated / predicted compensatory reserve can be based on a baseline estimate of the patient's compensatory reserve established when the patient has normal blood volume. In yet other cases, the estimation / prediction may not be based on a baseline estimate of the patient's compensatory reserve established when the patient has normal blood volume.

  [0129] By way of example only, FIG. 3B directly derives an assessment of the hydration effect (eg, before, during, and / or after resuscitation by infusion) and / or an assessment of blood loss directly from the sensor data. To derive, one technique 300 ′ is shown for deriving an estimate of CRI according to some embodiments similar to technique 200 ′ described above with respect to FIG. 2B (in fact, CRI is described herein). And the derived value can be used alone or with raw sensor data to evaluate such effectiveness). The illustrated techniques include waveform data (eg, including but not limited to arterial waveform data such as continuous PPG waveforms and / or continuous non-invasive blood pressure waveforms) for a specified period such as 32 heart rates. And any of the data described in related applications) (block 370). The sample is compared to a plurality of waveforms of reference data corresponding to different CRI values (block 375). (These reference waveforms may be derived using the algorithm described in the related application, or may be the result of experimental data or the like). By way of example only, the sample is compared to the waveform corresponding to the case where CRI is 1 (block 375a), CRI is 0.5 (block 375b), and CRI is 0 (block 375c), as shown. can do. From the comparison, a similarity factor is calculated (eg, using least squares or similar analysis) to represent the similarity between each of the sampled waveform and the reference waveform (block 380). These similarity factors can be normalized (if appropriate) (block 385) and summed (block 390) to produce an estimate of the patient's CRI (block). 395).

  [0130] Referring again to FIG. 3A, the method 300 may comprise estimating and / or predicting a patient's dehydration (block 330). A patient's dehydration can be expressed in a number of ways. For example, the state of dehydration can be expressed as a normalized value (eg, 1.0 corresponding to a fully hydrated state and 0.0 corresponding to a pathological dehydration state). In other cases, the dehydration state can be expressed as an infusion deficit or as the amount of infusion present in the patient's organ or using any other suitable metric.

  [0131] Many techniques can be used to model dehydration. By way of example only, the relationship between a patient's compensatory reserve and dehydration level can be modeled as described above (and as described in more detail below). Thus, in some embodiments, estimating a patient's dehydration state is based on estimating the patient's compensatory reserve (eg, CRI) and based on the estimate and a known relationship. Can be provided. Similarly, the predicted value of the compensatory reserve at some point in the future can be used to derive the predicted dehydration state at that point in the future. Other techniques can use parameters other than CRI to model the dehydration state.

  [0132] The method 300 may further comprise normalizing the analysis results, such as compensatory reserve, dehydration status, and / or bleeding probability, to name a few examples (block 335). By way of example only, a patient's estimated / predictive compensatory reserve is a reference normal blood volume value corresponding to normal fluid volume, a reference excess blood volume value corresponding to circulatory overload, and a reference minimum blood volume corresponding to cardiovascular collapse. Can be normalized to the value. Any value can be selected as the reference value. Merely by way of example, in some embodiments, the reference excess blood volume value is> 1, the reference normal blood volume value is 1, and the reference minimum blood volume value is 0. Alternatively, in other embodiments, the reference excess blood volume value can be defined as 1, the reference normal blood volume value can be defined as 0, and the reference minimum blood volume value in cardiovascular collapse can be defined as -1. As can be seen from these examples, various embodiments can use a number of different scales to normalize CRI and other estimation parameters.

  [0133] In one aspect, normalizing the data can be beneficial in the clinical setting while allowing the clinician to quickly make a qualitative determination of the patient's condition, while providing a raw estimate Interpretation of values / predicted values may require additional analysis. By way of example only, with respect to the estimation of a patient's compensatory reserve, the estimate is normalized to a baseline normal blood volume value corresponding to normal fluid volume and a baseline minimum blood volume value corresponding to cardiovascular collapse. can do. Also in this case, an arbitrary value can be selected as the reference value. For example, if the standard normal blood volume is defined as 1 and the standard minimum blood volume value is defined as 0, the normalized value is between 0.0 and 1.0, and the normal fluid volume and cardiovascular The physician can be quickly informed of the patient's position on the continuum during collapse. Similar normalization procedures can be performed for other estimated data (such as probabilities such as bleeding and / or dehydration).

  [0134] The method 300 may further comprise displaying the data on a display device (block 340). Such data may include an estimate and / or predictive value of the patient's compensatory reserve and / or an estimate and / or predictive value of the patient's dehydration status. Various techniques can be used to display such data. Merely by way of example, in some cases, displaying an estimate of a patient's compensatory reserve may comprise displaying a normalized estimate of the patient's compensatory reserve. Alternatively and / or additionally, displaying a normalized estimate of the patient's compensatory reserve is a normalized excess blood volume value, a normalized normal blood volume value, a normalized minimum blood volume value, and Graphical plots showing normalized estimates of compensatory reserves (for example, normalized excess blood volume values, normalized normal blood volume values, normalized minimum blood volume values) Displaying can be provided.

  [0135] In some cases, the method 300 repeats the operations of monitoring the patient's physiological data, analyzing the physiological data, and estimating (and / or predicting) the patient's compensatory reserves. Producing an estimated (and / or predicted) compensatory reserve can be provided. Thus, displaying an estimate (and / or prediction) of a patient's compensatory reserve is a new estimate of the compensatory reserve to display a plot of the estimated compensatory reserve over time. Updating the display of an estimate of the compensatory reserve to indicate (and / or predicted value) may be provided. Thus, a patient's compensatory reserve can be any desired interval (eg, after every beat), on demand, before resuscitation with fluids, during resuscitation with fluids, after resuscitation with fluids, etc. One or more combinations can be repeatedly estimated and / or predicted.

  [0136] In a further embodiment, the method 300 may comprise determining a probability that the patient is bleeding and / or displaying an indication of the probability that the patient is bleeding using a display device. (Block 345). For example, some embodiments can generate a model based on data that removes bodily fluids from the circulatory system (eg, LBNP, dehydration, etc.). Another embodiment can generate a model based on a known volume of blood donation (eg, 500 cc), voluntarily from a subject, eg, based on bodily fluids removed during blood donation. Based on this model, techniques similar to those described above can be used to monitor and analyze the patient's physiological data to estimate the probability that the patient is bleeding (eg, internally).

[0137] In some cases, the probability that a patient is bleeding can be used to adjust the patient's estimated CRI. Specifically, considering the bleeding probability expressed as Pr_Bleed at time t, the adjustment value of CRI can be expressed as follows.

CRI _Adjusted (t) = 1 − ((1−CRI (t)) × Pr_Bleed (t)) (Formula 13)

  [0138] In this relationship, the estimated CRI can be adjusted to produce a more accurate diagnosis of the patient's condition at a given time.

  [0139] The method 300 may comprise selecting a recommended treatment option for the patient at the computer system and / or displaying the recommended treatment option at the display (block 355). Recommended treatment options include, but are not limited to, optimization of the patient's hemodynamics, ventilator adjustment, intravenous fluid adjustment, blood or blood product transfusion to the patient, volume expansion agent infusion to the patient, It can be any of a number of treatment options, including changing medications administered to the patient, changing the patient's location, and surgical treatment.

  [0140] In a particular non-limiting example, the method 300 comprises controlling the operation of the hemodialysis device based at least in part on an estimate of the patient's compensatory reserve (block 360). it can. By way of example only, a computer system that performs monitoring and estimation functions may be configured to adjust the ultrafiltration rate of the hemodialysis device in response to the patient's estimated CRI value. In other embodiments, the computer system can provide instructions or suggestions to the hemodialyzer operator, such as instructions for manually adjusting the ultrafiltration rate or the like.

  [0141] In some embodiments, the method 300 may include assessing an individual's tolerance to blood loss, general volume loss, and / or dehydration (block 365). For example, such an embodiment lay down from a sitting position (e.g., from lying down to standing up, from standing up to lying down, from lying down to sitting down) Estimating the patient's CRI based on the condition (from standing to sitting and / or sitting to standing). Based on changes in the patient's CRI in response to these movements, the patient's sensitivity to blood loss, volume loss, and / or dehydration can be measured. In one aspect, this measurement can be performed using the CRI model generated as described above, the patient can be monitored using one or more sensors described above, and the subject can change position. The change in sensor output can be analyzed according to a model (eg, as described above) to assess an individual's tolerance to volume loss. Such monitoring and / or analysis can be performed in real time.

[0142] Referring again to FIG. 2, data for a model (eg, can include multiple submodels, such as a BL model for raw data and a BL model for CRI), collected directly by a sensor, such as a CRI Based on the analysis of the patient's blood loss (block 220) based on the analysis of the patient's physiological data against the model. Can be included. As described above, assessing blood loss may include estimated efficacy of hydration effort, BL, volume of fluid required for effective hydration, V, and / or probability P _f that a patient requires fluid, etc. Of estimating and / or predicting various values of.

  [0143] In some cases, the assessment of blood loss is based on the analysis of multiple measured (or derived) values of a particular physiological parameter (or parameters). Thus, in some cases, analysis of data on a continuous waveform may be performed either during or after measurement of the waveform by the sensor (or both), and assessment of blood loss may include hydration efforts and / or If resuscitation efforts by infusion continue, they can be renewed. In addition, the amount of infusion added to the patient's blood volume can be measured directly, and these direct measurements are fed back to the model to update the model (block 225), thereby improving the performance of the algorithms in the model. Improve (eg, by refining the weights given to different parameters with respect to estimated or predicted values). The updated model can then be used to continue to evaluate treatment (in emergency and / or future patients) as shown by the dashed line in FIG. 2A.

  [0144] In some cases, the method 200 includes displaying data indicating an assessment of the effectiveness of hydration (block 230). In some cases, the data may be displayed on a display of the sensor device (such as device 105 shown in FIG. 1B). Alternatively and / or additionally, the data may be displayed on a dedicated machine, such as a compensatory reserve monitor, or on a monitor of a general purpose computer system. Data can be displayed in alphanumeric, graphic, or both. FIGS. 6-8 described below illustrate some possible exemplary displays of blood loss and / or CRI assessment. There are many different ways in which data can be displayed, and the evaluation, estimate, or predicted value generated by method 200 can be displayed in any desired manner, according to various embodiments.

  [0145] In certain embodiments, the method 200 selects and / or displays treatment options for a patient (block 235) and / or controls a treatment device (block) based on an assessment of the patient's blood loss (block 235). 240). For example, the display can tell the clinician or the patient himself that the patient is losing blood (or has lost), should start or continue resuscitation treatment by infusion, drink, infuse, or ingest The estimated amount of infusion, the infusion rate in the case of IV infusion, the infusion rate for the IV pump or infuser, etc. can be indicated. Similarly, the system may be based on an assessment of the effectiveness of hydration, such as dispensing infusion fluid from an automatic dispenser, facilitating or adjusting IV pump or infusor flow, and / or adjusting IV infusion rate. It can be configured to control the operation of the device. As another example, certain embodiments can include a water bag (eg, a backpack-based hydration pack, such as that available from Camelbak Products LLC) or a water bottle, Can communicate with such dispensing devices and / or control the operation of such dispensing devices (eg, causing the device to dispense a particular volume of infusion and triggering an audible alarm to the device).

[0146] Further, in certain embodiments, the method 200 can include functionality to assist a clinician (or other personnel) to monitor hydration, infusion resuscitation, and / or blood volume status. . For example, in some cases, some measure of efficacy outside the normal range (eg, a P _f value above a certain threshold, a BL value below a threshold, etc.) may cause an audible alert, a message to a physician, to a patient Various alert conditions are set, such as the following message, update information automatically written on the patient chart. Such messaging can be realized by e-mail, text message, etc., and the sensor device or monitoring computer can be configured with, for example, an SMTP client or a text messaging client to perform such messaging. .

  [0147] In some cases, feedback and / or notifications may be sent to a third party regardless of whether an alarm condition has occurred. For example, the hydration monitor may monitor results (e.g., ratings, estimates, and Or any of the predicted values) can be configured to be transmitted. Examples include wireless alarms, wearable devices (eg, smart watches or glasses), or other personal devices of a patient, eg, for inclusion in health monitoring applications (eg, Bluetooth). Trademark), NFC, WiFi, etc.). In addition, and / or, for example, to allow a coach to remotely monitor a player's hydration to allow the parent to remotely monitor the hydration of the child (or the child is an older parent). Such information to a designated device or computer (e.g., any available), such as to enable and / or to enable an officer to remotely monitor a soldier's hydration (Via IP connection). In some cases (for example, in the case of a coach or superior), the application can aggregate results from multiple hydration monitors and enable the administrator to hydrate against a group of people (eg, in a dashboard-type configuration) Allows viewing of sex and / or blood loss (and / or any other data such as CRI, blood pressure, etc.). Such a display may use multiple (fuel gauge) displays, for example one (or more) for each person in the group, allowing the administrator to quickly check for abnormal results (For example, based on gauge color).

  [0148] Similarly, if an alarm condition is met for another physiological parameter (eg, blood pressure that can be estimated as described in the '171 application), the alarm is determined by the hydration effect of method 200. Can be triggered to determine if the first alarm condition is beneficial. If not, there may be automatic silencing of the original alarm condition, probably because everything is now good. More generally, assessment techniques provide information to each other or work in combination to inform each other about how to maintain optimal physiological stability (as described in the related application). (Including but not limited to) can be added to the ecosystem of surveillance algorithms.

  [0149] FIG. 4 illustrates a method 400 using such a self-learning predictive model (or machine learning) technique, according to some embodiments. In particular, the method 400 can be used to correlate physiological data received from a subject sensor with a measured physiological condition. More specifically, for various embodiments, the method 400 includes blood loss, water, and water from one or more of a variety of different physiological parameters, including but not limited to those described above and in related applications. Evaluate, predict, and / or estimate various physiological parameters such as sum efficacy or infusion resuscitation effort, estimated and / or predicted blood pressure, CRI, probability of patient bleeding, and / or patient dehydration Can be used to generate a model for

[0150] The method 400 includes D data signals s ₁ ,. . . , S _D (each of the data signals s are in each case input from one or many different physiological sensors) collect raw data measurements that can be used to derive a set Starting at block 405. Embodiments are not constrained by the type of measurements made at block 405 and can generally operate on any data set. For example, the data signal can be retrieved from a computer memory and / or provided from a sensor or other input device. As a specific example, the data signal can correspond to the output of the sensor described above (measuring the type of waveform data described above, such as continuous, non-invasive PPG data and / or blood pressure waveform data).

[0151] A set of K current or future results
Is assumed at block 415 (in this case, the result o is the probability that an infusion is needed, the volume of infusion needed for effective hydration or resuscitation by infusion, BL, CRI, dehydration, bleeding probability, etc. Past and / or future physiological state). The method uses derived data signals
And results
Is automatically generated. As used herein, “autonomous” means “no human intervention”.

[0152] As shown in block 420, which may be achieved by identifying the most predictive set of signals S _k, where, S _k is derived signal s ₁ for each result o _k ,. . . , S _D at least partly (and possibly all), where kε {1,. . . , K}. Stochastic prediction model
Is learned at block 425 where:
Are all k∈ {1,. . . A prediction of the result o _k derived from the model M _k to be used as an input value obtained from the signal set S _k for K}. Method 400 includes signals s ₁ ,. . . , S _D and the corresponding results o ₁ ,. . . , O from data containing exemplary values of _K , predictive models
Can be learned in stages (block 430). Once the data is available, the method 400 may use all kε {1,. . . , K} loop so that data is added in stages to the model for the same or different signal set S _k .

[0153] While the above description outlines general features of the method, additional features are noted. A linear model framework can be used to identify predictors for each new increment of data. In certain embodiments, a finite set of signal and result data
Is given, the linear model is all kε {1,. . . , K} can have a shape.
(Formula 14)
Where f _k is an arbitrary mapping from one input to one output, a ₀ , a ₁ ,. . . , _Ad are linear model coefficients. The framework used to derive the linear model coefficients is which signal s, s ₁ ,. . . , S _d are not predictive, and accordingly corresponding coefficients a ₀ , a ₁ ,. . . , It can be estimated either set a _d to zero. Using only predictors, the model
Build a predicted density model. A new predictive density model can be built for each new data increment.

  [0154] In some embodiments, a predictive model may be used to predict future results from previously analyzed data and / or to modify the predictive model if the data does not fit the predictive model A possible prediction system can be implemented. In some embodiments, the prediction system can make predictions and / or adapt the prediction model in real time. Furthermore, in some embodiments, the prediction system can not only create a prediction model, but also use a large data set to predict future results and to fit the prediction model.

  [0155] In some embodiments, a self-learning predictor may include a data input, a processor, and an output. The memory can include application software that can instruct the processor to perform predictions from input data based on the prediction model at run time. Any type of predictive model that manipulates any type of data can be used. In some embodiments, the predictive model can be implemented for a particular type of data. In some embodiments, upon receipt of the data, the prediction model can determine whether to grasp the data according to the prediction model. If the data is known, a prediction is made and an appropriate output is provided based on the prediction model. If you do not know the data at the time of reception, you can modify the model by adding data to the prediction model. In some embodiments, the apparatus can wait to determine the results of the specified data, in which case the prediction model can be modified accordingly. In some embodiments, if the data is captured by the prediction model and the output generated using the prediction model is not accurate, the data and results can be used to modify the prediction model. In some embodiments, the modification of the prediction model can be done in real time.

  [0156] Certain embodiments are described in accordance with the methodology described herein, using the tools and techniques described in the related applications, using the cardiac reserve monitor, wrist-worn sensor device described herein, And / or can perform the functions of a monitoring computer (in some embodiments, some or all of the functions can be combined in a single integrated device). These functions include assessment of resuscitation by patient infusion, assessment of patient hydration, estimation of current or future blood pressure and / or compensatory reserves of subjects (including but not limited to patients) and / or Including, but not limited to, predicting, estimating and / or determining the probability that a patient is bleeding (and / or internally) bleeding and / or bleeding and / or recommending treatment options for such conditions . Such tools and techniques include, among other things, systems (eg, computer systems, sensors, therapeutic devices, etc.), methods (eg, analysis methods for generating and / or using analysis models, diagnostics) described in related applications. Methods), and software programs described in this specification and related applications, which are incorporated herein by reference.

  [0157] FIG. 5 illustrates a method 500 for performing rapid detection of bleeding before, during, and after infusion resuscitation, according to various embodiments. In the embodiment of FIG. 5, the method 500 comprises, at block 505, estimating a patient's CRI before, during, and / or after resuscitation (eg, resuscitation by infusion, etc.). The estimation of the patient's CRI can be performed using, for example, the techniques described above with respect to FIGS. 3A and 3B, or using other techniques described above.

  [0158] At block 510, the method 500 may comprise recording the patient's CRI before, during, and / or after resuscitation. In some cases, the CRI may be recorded or stored in one or more of a data storage device that is part of the processing unit 145 and / or a memory device that is part of the monitoring computer 100 of FIG. The method 500 may calculate an average CRI (block 515) over a period of K seconds (K> 1) before, during, and / or after resuscitation, and before, during, and / or after resuscitation. Calculate CRI standard deviation or variance later over a period of K seconds (K> 1) (block 520) and CRI over a period of K seconds (K> 1) before, during, and / or after resuscitation Calculating the Pearson moment coefficient of the degree of skew (block 525) and calculating the rate of change of CRI over a period of K seconds (K> 1) before, during and / or after resuscitation (block 530) and the rate of CRI rate change (also referred to herein as “CRI acceleration”) over a period of K seconds (K> 1) before, during, and / or after resuscitation (ie, rate change) Rate Calculating a slip) (block 535) may further comprise a.

  [0159] According to some embodiments, the method 500 is based at block 540 on one or more of the calculations at blocks 515-535 (which may be referred to herein as "variation results"). The method may further comprise determining a probability of bleeding. In other words, the variability results are indicated by one or more bleeding states: (specific) non-bleeding status (probably indicated by the symbol “0”), (specific) bleeding state (probably indicated by the symbol “1”). ), And the probability of a bleeding condition (probably indicated between “0” and “1”).

  [0160] In some embodiments, (i) a CRI value sample, (ii) a set of CRI values, (iii) an average CRI, (iv) a median CRI, (v) a standard deviation of CRI, (vi) The following definitions can be used for CRI rate of change, (viii) rate of change of CRI rate change, and (viii) CRI skewness.

[0161] (i) Specific CRI value at time t:

CRI (t) (Formula 15)

[0162] (ii) Time {t ₁ , t ₂ ,. . . , T _K }, the set of CRI values:

CRI = {CRI (t ₁ ), CRI (t ₂ ),..., CRI (t _K )} (Equation 16)

[0163] (iii) Time {t ₁ , t ₂ ,. . . , T _K } average CRI values over a specific set:
(Formula 17)

[Iv] (iv) Time {t ₁ , t ₂ ,. . . , T _K } median CRI values over a specific set:
(Formula 18)

[0165] (v) Time {t ₁ , t ₂ ,. . . , T _K }, a measure of the deviation of the CRI over a particular set, possibly the variance defined by:
(Formula 19)

(Vi) A set of CRI values {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} is the rate of change of CRI expressed in m _K, which can be calculated as a slope of a line, measuring some increase or decrease in CRI over a particular period of time, for example. it can:
(Formula 20)

[0167] where A is a matrix defined by:
(Formula 21)

[Vii] A set of CRI values {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} is the rate of change of the rate of change of the CRI represented by r _K , and the rate of change of rate of change measures the rate of change of some increase or decrease in CRI over a specified period. For example, it can be calculated as a quadratic increase or decrease of the curve:
(Formula 22)

[0169] where B is a matrix defined by:
(Formula 23)

[V170] A set of CRI values {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} over several measurements of the skewness indicated by S _K (not to be confused with the signal set S _k , as described above with respect to FIG. 4), and S _K is probably the Fisher Pearson's skewness coefficient variant:
(Formula 24)

[0171] and / or _{S K} is skewness is defined as follows, is probably another some other measure of skewness (or skewness of Bowley) of Galton:
(Formula 25)

[0172] The method of estimating the (specific) non-bleeding state is probably, but not limited to, one or more of the following calculations, perhaps within a statistical and / or machine learning framework: A combination of such calculations of: (1) the average value of CRI before resuscitation ("
")> NB1, (2) Average CRI during resuscitation ("
")> NB2, (3) Average CRI after resuscitation ("
")> NB3, (4)
−
> NB4, (5)
−
> NB5, (6)
−
> NB6, (7) Standard deviation or variance of CRI before resuscitation ("[SD (CRI)] _BR ") <NB7, (8) Standard deviation or variance of CRI during resuscitation ("[SD (CRI)] _DR" ") <NB8, (9) the standard deviation or variance of the CRI after resuscitation _{(" [SD (CRI)]} AR ") <NB9, (10) [SD (CRI)] AR - [SD (CRI)] BR < NB10, (11) moment coefficient of CRI skewness before resuscitation (positive or negative) (“S _BR ”) <NB11, (12) moment coefficient of skewness of CRI during resuscitation (positive or negative) (“S _{DR ") <NB12, (13} ) the moment coefficient of CRI skewness after resuscitation (positive or _{negative) (" S AR ") <} NB13, (14) CRI rate of change before resuscitation _{(" m} BR _")> NB14, (15) CRI change rate during resuscitation _{"M DR")> NB15,} (16) CRI of variation after resuscitation _{( "m AR")> NB16} , (17) m AR -m BR> NB17, (18) m DR -m BR> NB18, ( 19) CRI rate change rate before resuscitation (“r _BR ”)> NB19, (20) CRI rate change rate during resuscitation (“r _DR ”)> NB20, (21) CRI rate change rate after resuscitation _{( "r AR")> NB21} , , and the like _{(22) r AR -r BR>} NB22, and / or _{(23) r DR -r BR>} NB23. In some cases, each or one or more of NB1 to NB23 may be estimated experimentally or set by the user. Here, the number K> 0 differs at each stage of the calculation from (1) to (23), and may be selected by the user or may be determined experimentally.

[0173] Referring to (1), average CRI before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation,
Can be the average of those points. So, for example,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold indicated by (for example, NB1 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 26)

[0174] Referring to (2), the average CRI during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation,
Can be the average of those points. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold indicated by (for example, NB2 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 27)

[0175] Referring to (3), the average CRI after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation,
Can be the average of those points. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold indicated by (for example, NB3 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 28)

[0176] Referring to (4),
and
Can be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, NB4 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 29)

[0177] Referring to (5)
and
Can be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, NB5 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 30)

[0178] Regarding (6),
and
Can be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, NB6 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 31)

[0179] Referring to (7), CRI dispersion before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and [SD (CRI)] _BR is the variance of those values (probably defined above) Standard deviation). Thus, for example, ^NB [SD (CRI)] _BR (eg, NB7 above) can be used to classify non-bleeding by selecting an experimental or user-set threshold, and non-bleeding classification Can be determined by the following equation.

[SD (CRI)] _BR < ^NB [SD (CRI)] _BR (Formula 32)

[0180] Referring to (8), CRI dispersion during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and [SD (CRI)] _DR is the variance of those values (perhaps as defined above) Standard deviation). Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold, indicated by ^NB [SD (CRI)] _DR (eg, NB8 above), It can be determined by the following formula.

[SD (CRI)] _DR < ^NB [SD (CRI)] _DR (Formula 33)

[0181] With respect to (9), CRI dispersion after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation and [SD (CRI)] _AR is the variance of those values (perhaps as defined above) Standard deviation). Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold, indicated by ^NB [SD (CRI)] _AR (eg, NB9 above), It can be determined by the following formula.

[SD (CRI)] _AR < ^NB [SD (CRI)] _AR (Formula 34)

[0182] Referring to (10), [SD (CRI)] _BR and [SD (CRI)] _AR may be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold indicated by (for example, NB10 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 35)

[0183] Referring to (11), CRI skewness before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and S _BR is the skewness of those points (perhaps as defined above). It can be a measured value. Thus, for example, non-bleeding classification can be performed by selecting an experimental or user-set threshold, indicated by ^NB S _BR (e.g., NB11 above). It can be determined by whether or not.

| S _BR | < ^NB S _BR (Formula 36)

[0184] Referring to (12), CRI skewness during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and _SDR is a measure of the skewness of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^NB S _DR (for example, NB12 above). Can be determined.

| S _DR | < ^NB S _DR (Formula 37)

[0185] For (13), CRI skewness after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation and _SAR is a measure of the skewness of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^NB S _AR (for example, NB13 above). Can be determined.

| S _AR | < ^NB S _AR (Formula 38)

Referring to (14), the rate of change of CRI before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and m _BR is the rate of change of those points (perhaps as defined above). It can be a measured value. Thus, for example, non-bleeding classification can be performed by selecting an experimental or user-set threshold, indicated by ^NB m _BR (e.g., NB14 above). It can be determined by whether or not.

m _BR > ^NB m _BR (Formula 39)

[0187] Referring to (15), the rate of change of CRI during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and m _DR is a measure of the rate of change of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^NB m _DR (for example, NB15 above). Can be determined.

m _DR > ^NB m _DR (Formula 40)

[0188] For (16), the rate of change of CRI after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation, and m _AR is a measure of the rate of change of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^NB m _AR (for example, NB16 above). Can be determined.

m _AR > ^NB m _AR (Formula 41)

[0189] Referring to (17), m _BR and m _AR may be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, NB17 above), and the non-bleeding classification can be determined by the following equation: it can.
(Formula 42)

[0190] Referring to (18), m _BR and m _DR may be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (e.g., NB18 above), and the non-bleeding classification can be determined by the following equation: it can.
(Formula 43)

[0191] Referring to (19), the rate of change in CRI ratio before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and r _BR is the rate of rate change of those points (perhaps as defined above). Measured value. Thus, for example, a non-bleeding classification can be performed by selecting an experimental or user-set threshold, as indicated by ^NB r _BR (eg, NB19 above). It can be determined by whether or not.

r _BR > ^NB r _BR (Formula 44)

[0192] Referring to (20), CRI rate change rate during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and r _DR is the percentage change rate of those points (perhaps as defined above). It can be a measured value. Therefore, the non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^NB r _DR (for example, NB20 above). Can be determined.

r _DR > ^NB r _DR (Formula 45)

[0193] With regard to (21), the rate of change of CRI after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time after resuscitation, and r _AR is the percentage change in those points (perhaps as defined above). It can be a measured value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^NB r _AR (for example, NB21 above). Can be determined.

r _AR > ^NB r _AR (Equation 46)

[0194] Referring to (22), r _BR and r _AR may be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, NB22 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 47)

[0195] Referring to (23), m _BR and m _DR may be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold indicated by (for example, NB23 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 48)

[0196] Similarly, in some cases, the method of estimating a (specific) bleeding state is not limited, but perhaps of the following calculations, such as within a statistical and / or machine learning framework: One or more combinations of two or more such calculations may be included: (1) the average value of CRI before resuscitation (“
") <BL1, (2) Average CRI during resuscitation ("
") <BL2, (3) Average CRI after resuscitation ("
") <BL3, (4)
−
<BL4, (5)
−
<BL5, (6)
−
<BL6, (7) Standard deviation of CRI before resuscitation ("[SD (CRI)] _BR ")> BL7, (8) Standard deviation of CRI during resuscitation ("[SD (CRI)] _DR ")> BL8 (9) Standard deviation of CRI after resuscitation (“[SD (CRI)] _AR ”)> BL9, (10) [SD (CRI)] _AR− [SD (CRI)] _BR > BL10, (11) Resuscitation Moment factor of skewness of previous CRI (positive or negative) (“S _BR ”)> BL11, (12) Moment factor of skewness of CRI during resuscitation (positive or negative) (“S _DR ”)> BL12, (13) Moment coefficient of CRI skewness after resuscitation (positive or negative) (“S _AR ”)> BL13, (14) CRI change rate before resuscitation (“m _BR ”) <BL14, (15) Resuscitation CRI of the rate of change in ₍ "m DR") <BL 5, (16) CRI of variation after resuscitation _{( "m AR") <BL16} , (17) m AR -m BR <BL17, (18) m DR -m BR <BL18, (19) prior to resuscitation CRI Rate change rate (“r _BR ”) <BL19, (20) rate change rate of CRI during resuscitation (“r _DR ”) <BL20, (21) rate change rate of CRI after resuscitation (“r _AR ”) _<BL21, and so forth _{(22) r AR -r BR <} BL22, and / or _{(23) r DR -r BR <} BL23. In some cases, each or one or more of BL1 to BL20 may be estimated experimentally or set by the user.

[0197] Referring to (1), mean CRI before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation,
Can be the average of those points. So, for example,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold value indicated by (for example, BL1 above), and the non-bleeding classification can be determined by the following formula: it can.
(Formula 49)

[0198] Referring to (2), the average CRI during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation,
Can be the average of those points. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL2 above), and the non-bleeding classification can be determined by the following equation: it can.
(Formula 50)

[0199] Referring to (3), the average CRI after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation,
Can be the average of those points. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL3 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 51)

[0200] Referring to (4),
and
Can be as defined above. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL4 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 52)

[0201] Referring to (5)
and
Can be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, BL5 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 53)

[0202] Regarding (6),
and
Can be as defined above. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL6 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 54)

[0203] Referring to (7), CRI dispersion before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and [SD (CRI)] _BR is the variance of those values (probably defined above) Standard deviation). Thus, for example, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B [SD (CRI)] _BR (eg, BL7 above). Can be determined by the following equation.

[SD (CRI)] _BR > ^B [SD (CRI)] _BR (Formula 55)

[0204] Referring to (8), CRI dispersion during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and [SD (CRI)] _DR is the variance of those values (perhaps as defined above) Standard deviation). Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B [SD (CRI)] _DR (eg, BL8 above), It can be determined by the following formula.

[SD (CRI)] _DR > ^B [SD (CRI)] _DR (Formula 56)

[0205] With regard to (9), CRI dispersion after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation and [SD (CRI)] _AR is the variance of those values (perhaps as defined above) Standard deviation). Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B [SD (CRI)] _AR (eg, BL9 above), It can be determined by the following formula.

[SD (CRI)] _AR > ^B [SD (CRI)] _AR (Formula 57)

[0206] Referring to (10), [SD (CRI)] _BR and [SD (CRI)] _AR may be as defined above. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL10 above), and the non-bleeding classification can be determined by the following formula it can.
(Formula 58)

[0207] Referring to (11), the skewness of CRI before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and S _BR is the skewness of those points (perhaps as defined above). It can be a measured value. Thus, for example, non-bleeding classification can be performed by selecting an experimental or user-set threshold shown by ^B S _BR (eg, BL11 above). It can be determined by whether or not.

| S _BR |> ^B S _BR (Formula 59)

[0208] Referring to (12), CRI skewness during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and _SDR is a measure of the skewness of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B S _DR (eg, BL12 above). Can be determined.

| S _DR |> ^B S _DR (Formula 60)

[0209] With regard to (13), CRI skewness after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation and _SAR is a measure of the skewness of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B S _AR (for example, BL13 above). Can be determined.

| S _AR |> ^B S _AR (Formula 61)

[0210] Referring to (14), the rate of change of CRI before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and m _BR is the rate of change of those points (perhaps as defined above). It can be a measured value. Thus, for example, non-bleeding classification can be performed by selecting an experimental or user-set threshold, as indicated by ^B m _BR (eg, BL14 above). It can be determined by whether or not.

m _BR < ^B m _BR (Formula 62)

[0211] Referring to (15), the rate of change of CRI during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and m _DR is a measure of the rate of change of those points (possibly as defined above). Can be a value. Therefore, classification of non-bleeding can be performed by selecting an experimental or user-set threshold indicated by ^B m _DR (for example, BL15 above). Can be determined.

m _DR < ^B m _DR (Formula 63)

[0212] For (16), the rate of change of CRI after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled after resuscitation, and m _AR is a measure of the rate of change of those points (possibly as defined above). Can be a value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B m _AR (eg, BL16 above). Can be determined.

m _AR < ^B m _AR (Formula 64)

[0213] Referring to (17), m _BR and m _AR may be as defined above. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL17 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 65)

[0214] Referring to (18), _mBR and _mDR may be as defined above. Therefore,
The classification of non-bleeding can be performed by selecting an experimental or user-set threshold shown in (for example, BL18 above), and the classification of non-bleeding can be determined by the following equation it can.
(Formula 66)

[0215] Referring to (19), the rate of change of CRI ratio before resuscitation, CRI _BR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time prior to resuscitation, and r _BR is the rate of rate change of those points (perhaps as defined above). Measured value. Thus, for example, ^{B r} BR _(e.g., above BL19) represented by the empirical or user can perform the classification of non-bleeding by selecting a threshold set, the classification of non-bleeding, the following formula It can be determined by whether or not.

r _BR < ^B r _BR (Formula 67)

[0216] Referring to (20), CRI rate change rate during resuscitation, CRI _DR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at the time of resuscitation, and r _DR is the percentage change rate of those points (perhaps as defined above). It can be a measured value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B r _DR (eg, BL20 above). Can be determined.

r _DR < ^B r _DR (Formula 68)

[0217] With regard to (21), the rate of change of CRI after resuscitation, CRI _AR = {CRI (t ₁ ), CRI (t ₂ ),. . . , CRI (t _K )} can be any set of points sampled at a time after resuscitation, and r _AR is the percentage change in those points (perhaps as defined above). It can be a measured value. Therefore, non-bleeding classification can be performed by selecting an experimental or user-set threshold indicated by ^B r _AR (for example, BL21 above). Can be determined.

r _AR < ^B r _AR (Formula 69)

[0218] Referring to (22), r _BR and r _AR may be as defined above. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL22 above), and the non-bleeding classification can be determined by the following equation it can.
(Formula 70)

[0219] Referring to (23), m _BR and m _DR may be as defined above. Therefore,
Non-bleeding classification can be performed by selecting an experimental or user-set threshold shown in (for example, BL23 above), and the non-bleeding classification can be determined by the following equation: it can.
(Formula 71)

[0220] Similarly, in some cases, the method of estimating the probability of bleeding (eg, indicated by a numerical value between 0 and 1) is not limited, but probably to estimate the probability of bleeding Can include one of the above calculations or a combination of two or more such calculations, such as in a statistical and / or machine learning framework. In some cases, the method may include, but is not limited to, empirical estimation of probability density functions, and / or cumulative distribution functions using graphical and / or nonparametric models. Other methods may include, but are not limited to: (i) the probability of bleeding is proportional to the number of times the bleeding threshold is reached, (ii) the probability of not bleeding is the non-bleeding threshold (Iii) The probability of bleeding is proportional to the number of times that the non-bleeding threshold has been subtracted from the number of times that the bleeding threshold has been reached, (iv) The bleeding probability is Represented.
(Formula 72)

  [0221] where f is an empirical estimate of the probability density function and / or cumulative distribution function using a graphical model and / or a nonparametric model.

  [0222] In some embodiments, estimated CRI values include, but are not limited to, CRI values estimated or measured for each heartbeat, over the previous or last N seconds (N> 1). One or more can be included, such as an averaged CRI value and / or a median CRI over the previous or last N seconds (where N> 1). According to some embodiments, the calculations described above with respect to blocks 515 to 535 can utilize these estimated CRI values. According to some embodiments, instead of using a CRI measurement, the method uses the CRI value as an optional measure of compensatory reserve, or one or more thereof, using one or more of the sensor types described above. Any or all of the above calculations can be used that replace the values corresponding to the measurements associated with the derivation.

[0223] FIGS. 6-8 illustrate exemplary screen captures from a display of a compensatory reserve monitor showing various features that can be provided by one or more embodiments. Similar screens may be shown on other monitoring devices such as a wrist-mounted sensor device display and / or a monitoring computer display. While FIGS. 6-8 use BL or CRI as exemplary conditions for exemplary purposes, other embodiments provide values for volume V, the volume of infusion required for effective hydration, or The probability P _f that the patient needs an infusion (including additional infusion if a hydration effort is already in progress) can be displayed.

  [0224] FIG. 6 shows a compensatory reserve monitor implementation where a normalized CRI estimate of “1” means that blood loss is certain and “0” means no blood loss. An exemplary display 600 is shown. A value between “0” and “1” means a series of blood loss probabilities.

  [0225] FIG. 7A shows BL as a “fuel gauge” type bar graph for a person undergoing central volume blood loss and subsequent hydration efforts, or who is undergoing, is receiving or has undergone infusion resuscitation. FIG. 6 shows four screen captures 700 of a display of a compensatory reserve monitor implementation for display. 6 shows a CRI trace over time, the bar graph of FIG. 7A provides a snapshot of the BL at each screen capture corresponding to the CRI of FIG. (In the illustrated embodiment, the bar graph is updated continuously and / or periodically so that each bar graph can correspond to a particular position on the X-axis of FIG. 6).

  [0226] Various additional features are possible. By way of example only, FIG. 7B shows a similar “fuel gauge” type display, but the display is, for example, green (indicated by hatched hatching), yellow (in check pattern) corresponding to various levels of CRI. And red (shown in gray shading) and various color bars, as well as arrows 710 indicating CRI value trends, which indicate that blood loss has occurred and / or resuscitation efforts are active Indicates that

  [0227] In some embodiments, such "fuel gauge" indications (or other indicators of BL or CRI and / or different physiological parameters) can be incorporated into a more comprehensive user interface. By way of example only, FIG. 8 shows an exemplary display 800 of the monitoring system. The display 800 includes a graphical, color-coded “fuel gauge” type display 805 of the current estimate BL (similar to the display shown in FIG. 8B), along with a history display 810 of the latest CRI estimates, in this example. , Each bar in the history display 810 corresponds to an estimate run every minute, but different estimation frequencies are possible, and in some embodiments, the operator can be given the option of specifying various frequencies . In the illustrated embodiment, display 800 also includes a current BL numerical display 815 and a trend indicator 820 (similar to that described above).

  [0228] In certain embodiments, display 800 can include additional information (in some cases, the type of information displayed and / or the type of display can be configured by the operator). For example, exemplary display 800 includes a patient's current heart rate indicator 825 and a patient's blood oxygen saturation level (SpO 2) indicator 830. The exemplary display 800 also includes an indicator of the estimated amount V required for effective hydration, a numeric indicator 840, a trend indicator 845, and a similar color-coded “fuel gauge” display 850 of the current CRI. . Other monitoring parameters may be displayed as well, such as ECG traces, blood pressure, and / or probability of bleeding estimates.

  [0229] Exemplary clinical studies

  [0230] FIGS. 9A-9H (collectively, FIG. 9) illustrate prior to resuscitation by patient infusion in a multi-organ traumatic clinical trial at the Denver Health Medical Center (DHMC), according to various embodiments; FIG. 9 is a graphical diagram 900 showing rapid detection of bleeding during and after. In one exemplary multitraumatic clinical study in DHMC, 50 patients were enrolled, 45 of whom met the required criteria and 5 were excluded (having incomplete data and / or equipment ). Of 45 patients, 12 were bleeding (initial CRI value 0.17 ± 0.07, mean injury severity score (ISS) 27 ± 12.7), 30 were non-bleeding (initial CRI value 0.56 ± 0.17, mean ISS 7.5 ± 8.7), and 3 undefined.

  [0231] Referring to FIG. 9, FIG. 9A shows a receiver operating characteristic (ROC) curve used to classify bleeding using compensatory reserves. The sensitivity is 0.93, the specificity is 0.92, and the area under the curve (AUC) is 0.97.

  [0232] FIG. 9B shows the CRI for non-bleeding patients (shown in the graph as “trauma-free bleeding”) and bleeding patients (shown in the graph as “trauma + bleeding”). As shown in FIG. 9B, the CRI value during bleeding is low.

  [0233] FIGS. 9C-9E show line traces of actual CRI curves of three representative patients in the non-bleeding group. The CRI value for non-bleeding patients before infusion of intravenous fluid (IVF) is 0.56 ± 0.17. FIG. 9C shows the CRI curve of non-bleeding trauma patient 003 with a CRI of> 0.3 prior to IVF infusion with IVF containing 2 L of saline. There was no persistent decline in the CRI of this patient during or after IVF infusion. FIG. 9D shows the CRI curve of non-bleeding trauma patient 042 with a CRI of 0.4 prior to IVF infusion with IVF containing 1 L of saline. During or after IVF infusion, the patient's CRI had no wounds or persistent decline. FIG. 9E shows the CRI curve of a non-bleeding trauma patient 018 with a CRI of 0.65 prior to infusion of 2 L saline, IVF, where IVF is 1 L lactate Ringer (LR) solution and 2 packs Contains red blood cells (PRBC). There was no persistent decline in the CRI of this patient during or after IVF infusion. As shown in FIGS. 9C to 9E, CRI increases or generally increases during and after resuscitation with non-bleeding group infusions.

  [0234] FIGS. 9F through 9H show line traces of actual CRI curves of three representative patients in the bleeding group. The CRI value for non-bleeding patients before infusion of intravenous fluid (IVF) is 0.17 ± 0.07. FIG. 9F shows the CRI curve of a bleeding trauma patient 019 with a CRI of 0.15 prior to IVF injection (time 905) and with a first IVF injection (time 910), where the first IVF is 7 L saline, 3 packets of PRBC, 1 packet of platelets (PLT), and 3 packets of fresh frozen plasma (FFP). CRI decreased after the first increase (as shown at time 915). At time 920, a second IVF is infused, and the second IVF contains 4 L saline, 3 packets of PRBC, and 3 packets of fresh frozen plasma (FFP). FIG. 9G shows the CRI curve of bleeding trauma patient 006 with a CRI of 0.15 prior to IVF infusion (time 925) and with a first IVF infusion (time 930), where the first IVF is Contains 2 L of saline. CRI decreased after the first increase (as shown at time 935). At time 940, a second IVF is infused and the second IVF contains 1 L of saline. Again, the CRI decreased (as shown at time 945). FIG. 9H shows the CRI curve of bleeding trauma patient 012 with a first IVF (time 950) and second IVF (time 955) infusion with a CRI of 0.15 prior to IVF infusion, One IVF contains 1 L of saline and the second IVF contains 2.25 L of saline. As shown in FIGS. 9F-9H, CRI decreases after the first increase (during and after resuscitation with fluids) for the bleeding group.

  [0235] Exemplary System and Hardware Implementation

  [0236] FIG. 10 is a block diagram illustrating an exemplary computer or system hardware architecture, according to various embodiments. FIG. 10 may perform a method provided by various other embodiments as described herein, and / or a monitoring computer, CRI monitor, and / or sensor device, as described above. FIG. 2 provides a schematic diagram of an embodiment of a computer system 1000 that can function as a processing unit of Note that FIG. 10 is meant only to provide a generalized view of the various components, one or more (or none) of which can be utilized as needed. I want. Accordingly, FIG. 10 broadly illustrates how individual system elements can be implemented in a relatively isolated or relatively more integrated manner.

  [0237] Computer or hardware system 1000 is shown with hardware elements that can be electrically coupled (or communicated as needed) via bus 1005. One hardware element includes, but is not limited to, one or more general purpose processors and / or one or more application specific processors (such as a digital signal processing chip and / or a graphics accelerator processor). Or may include, but is not limited to, one or more input devices 1015, display devices and / or printers, which may include, but are not limited to, multiple processors 1010, mice and / or keyboards, etc. One or more output devices 1020 can be included, but not limited to.

  [0238] The computer or hardware system 1000 can include, but is not limited to, local and / or network accessible storage, and / or includes, but is not limited to, disk drives, drive arrays, optical One or more storages that can be programmable and / or flash updatable, which can include solid state storage devices such as storage devices, random access memory (RAM) and / or read only memory (ROM) A device 1025 can further be included (and / or communicated). Such a storage device can be configured to implement any suitable data store including, but not limited to, various file systems and / or database structures and the like.

  [0239] A computer or hardware system 1000 may also include, but is not limited to, a modem, a network card (wireless or wired), an infrared communication device, a wireless communication device, and / or a chipset (eg, a Bluetooth device). , 802.11 devices, WiFi devices, WiMax devices, WWAN devices, cellular communication equipment, etc.). Communication subsystem 1030 enables data exchange with a network (for example, the network described below), other computers or hardware systems, and / or any other device described herein. can do. In many embodiments, the computer or hardware system 1000 further comprises a working memory 1035 that may include a RAM or ROM device, as described above.

  [0240] A computer or hardware system 1000 is also provided by an operating system 1040, device drivers, executable libraries, and / or various embodiments (including, but not limited to, hypervisors and VMs). Software elements shown as currently located in working memory 1035, including other code such as one or more application programs 1045 that can provide a computer program, and / or Can be designed to implement methods and / or configure such systems as provided by other embodiments, as described in. Merely by way of example, one or more of the procedures described with respect to the methods described above may be implemented as code and / or instructions executable by a computer (and / or a processor within the computer), and in one aspect such Using code and / or instructions, a general purpose computer (or other device) can be configured and / or adapted to perform one or more operations in accordance with the described methods.

  [0241] These sets of instructions and / or codes may be encoded and / or stored on a non-transitory computer readable storage medium, such as the storage device (s) 1025 described above. In some cases, the storage medium may be incorporated within a computer system such as system 1000. In other embodiments, the storage medium may be separate from the computer system (ie, a removable medium such as a compact disk) and / or provided in an installation package, The general purpose computer can be used to program, configure, and / or adapt using instructions / code stored in the storage medium. These instructions can take the form of executable code executable by a computer or hardware system 1000 and / or (eg, various commonly available compilers, installation programs, compression / decompression utilities, etc.). Can be in the form of source and / or installable code that later takes the form of executable code when compiled and / or installed on the computer or hardware system 1000.

  [0242] It will be apparent to those skilled in the art that substantial changes can be made according to specific requirements. For example, customized hardware (such as programmable logic controllers, field programmable gate arrays, and / or application specific integrated circuits) may be used, and / or certain elements may be hardware, software (portable, such as applets) Software), or both. In addition, connections to other computing devices such as network input / output devices may be utilized.

  [0243] As described above, in one aspect, some embodiments use a computer or hardware system (such as a computer or hardware system 1000) to perform the methods according to various embodiments of the invention. be able to. According to one set of embodiments, some or all of such method steps may be incorporated into one or more instructions (operating system 1040 and / or other code) stored in working memory 1035. Executed by a computer or hardware system 1000 in response to a processor 1010 executing one or more sequences of application programs 1045, etc.). Such instructions may be read into working memory 1035 from another computer readable medium, such as one or more storage devices 1025. Merely by way of example, execution of a sequence of instructions contained in working memory 1035 may cause processor 1010 to perform one or more procedures of the methods described herein.

  [0244] The terms "machine-readable medium" and "computer-readable medium" as used herein refer to any medium that participates in providing data that causes a machine to operation in a specific fashion. In an embodiment implemented using a computer or hardware system 1000, various computer-readable media may be involved in providing instructions / code for execution by processor 1010 and / or its Such instructions / codes can be used to store and / or carry (eg as signals). In many implementations, the computer-readable medium is a non-transitory, physical, and / or tangible storage medium. In some embodiments, computer readable media may take many forms, including but not limited to, non-volatile media or volatile media. Non-volatile media includes, for example, optical disks such as storage device 1025 and / or magnetic disks. Volatile media includes, but is not limited to, dynamic memory such as working memory 1035. In some alternative embodiments, the computer-readable medium includes coaxial cables, conductors, and optical fibers, including wires comprising bus 1005, and various components of communication subsystem 1030 (and / or other communication subsystem 1030). Medium for communicating with other devices) may take the form of transmission media including, but not limited to. In an alternative set of embodiments, the transmission medium takes the form of waves (including but not limited to radio, acoustic, and / or light waves, such as those generated during radio wave and infrared data communications). You can also.

  [0245] Common forms of physical and / or tangible computer readable media are, for example, floppy disks, flexible disks, hard disks, magnetic tapes, or any other magnetic medium, CD-ROM, any other optical Media, punch card, paper tape, any other physical medium with a hole pattern, RAM, PROM, and EPROM, FLASH-EPROM, any other memory chip or cartridge, carrier wave described below, or computer command and / or Or any other medium from which the code can be read.

  [0246] Various forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to processor 1010 for execution. Merely by way of example, the instructions may initially be carried on a remote computer magnetic disk and / or optical disk. The remote computer can load the instructions into its dynamic memory and send the instructions as a signal over a transmission medium for reception and / or execution by the computer or hardware system 1000. These signals, which can be in the form of electromagnetic signals, acoustic signals, and / or optical signals, are all examples of carrier waves on which instructions can be encoded, according to various embodiments of the present invention.

  [0247] Communication subsystem 1030 (and / or its components) generally receives a signal, and then bus 1005 carries the signal (and / or data carried by the signal, instructions, etc.) to working memory 1035. From the working memory 1035, the processor 1005 retrieves and executes the instructions. The instructions received by working memory 1035 may optionally be stored on storage device 1025 either before or after execution by processor 1010.

  [0248] Conclusion

  [0249] This document discloses novel tools and techniques related to patient blood loss, compensatory reserves, and similar physiological conditions (eg, before, during, and / or after resuscitation by infusion). While particular features and aspects have been described with respect to exemplary embodiments, those skilled in the art will recognize that many variations are possible. For example, the methods and processes described herein can be implemented using hardware components, software components, and / or any combination thereof. Furthermore, for ease of explanation, the various methods and processes described herein can be described with respect to particular structural and / or functional components, but the methods provided by the various embodiments include: It is not limited to any particular structure and / or functional architecture, and can be implemented with any suitable hardware, firmware, and / or software configuration. Similarly, a particular function is attributed to a particular system component, but unless otherwise indicated, this functionality can be distributed among various other system components according to some embodiments.

  [0250] Further, the methods and process procedures described herein are set forth in a specific order for ease of explanation, but according to various embodiments, unless otherwise indicated by context. Various procedures can be rearranged, added, and / or omitted. Further, procedures described with respect to one method or process may be incorporated into other described methods or processes, as well as system components described in accordance with a particular structural architecture and / or with respect to a system. May be organized in alternative structural architectures and / or incorporated within other described systems. Accordingly, for ease of explanation and to illustrate exemplary aspects of those embodiments, various embodiments have been described with (or without) specific features, but specific Various components and / or features described herein with respect to the embodiments may be substituted, added, and / or subtracted from other described embodiments unless the context clearly dictates otherwise. it can. Thus, although several exemplary embodiments have been described above, it will be understood that the invention is intended to cover all modifications and equivalents within the scope of the appended claims.

Claims (35)

A system,
One or more sensors for obtaining physiological data from the patient;
A computer system in communication with the one or more sensors, the computer system comprising:
One or more processors;
A non-transitory computer readable medium in communication with the one or more processors, the computer system comprising:
Receiving the physiological data from the one or more sensors;
Analyzing the physiological data;
Estimating the probability that the patient is bleeding,
On top of which is an instruction set executable by the one or more processors for causing the display device to display at least one of an evaluation, a prediction value, or an estimate indicating the probability that the patient is bleeding. A non-transitory computer readable medium encoded with:
A system comprising: A method,
Monitoring patient physiological data using one or more sensors;
Analyzing the physiological data in a computer system;
Estimating the probability that the patient is bleeding based on at least a portion of the analyzed physiological data;
Displaying on the display device at least one indication of an estimate, a predicted value, or an estimated value of the probability that the patient is bleeding;
A method comprising:   Monitoring the physiological data; analyzing the physiological data; estimating the probability that the patient is bleeding; or evaluating, predicting or estimating the probability that the patient is bleeding The method of claim 2, wherein one or more of displaying at least one indication of a value is performed in real time.   Estimating the probability that the patient is bleeding may include determining whether the patient is bleeding based on one or more values of compensatory reserve index (CRI) values estimated based on the received physiological data. The method of claim 2, comprising estimating a probability of doing so.   The one or more values of CRI are estimated based on physiological data that is at least one of received before, during, or after a resuscitation procedure with an infusion. Item 5. The method according to Item 4.   The one or more values of CRI comprise a plurality of CRI values, and estimating the probability that the patient is bleeding is based at least in part on an average value of the CRI over a specific period of time 5. The method of claim 4, comprising estimating the probability that a person is bleeding.   The one or more values of CRI comprise a plurality of CRI values, and estimating the probability that the patient is bleeding is based at least in part on at least some standard deviations of the plurality of CRI values 5. The method of claim 4, comprising estimating the probability that the patient is bleeding.   The one or more values of CRI comprise a plurality of CRI values, and estimating the probability that the patient is bleeding is based at least in part on at least some skewness of the plurality of CRI values 5. The method of claim 4, comprising estimating the probability that the patient is bleeding.   The one or more values of CRI comprise a plurality of CRI values, and estimating the probability that the patient is bleeding is based at least in part on a rate of change of at least some of the plurality of CRI values 5. The method of claim 4, comprising estimating the probability that the patient is bleeding.   The one or more values of CRI comprise a plurality of CRI values, and estimating the probability that the patient is bleeding is at least partially in at least some rate of rate change of the plurality of CRI values. 5. The method of claim 4, comprising estimating the probability that the patient is bleeding based.   The one or more values of CRI comprise a plurality of CRI values, and estimating the probability that the patient is bleeding is based at least in part on a difference between at least some of the plurality of CRI values. 5. The method of claim 4, comprising estimating the probability that the patient is bleeding.   The method of claim 4, wherein the indication is a value between 0 and 1.   The method of claim 12, wherein a value of 0 indicates that the patient is not bleeding and a value of 1 indicates that the patient is bleeding. Estimating the patient's CRI comprises estimating a compensatory reserve index by comparing the physiological data to a model constructed using the following equation:
Where CRI (t) is the compensatory reserve at time t, BLV (t) is the subject's intravascular volume loss at time t, and BLV _HDD is the hemodynamic decompensation of the subject Intravascular volume loss in terms of
The method of claim 4.   The physiological data comprises waveform data, and the step of estimating the CRI is directed to a state of hemodynamic decompensation or near hemodynamic decompensation, or toward hemodynamic decompensation. The method of claim 4, comprising comparing the waveform data with one or more sample waveforms generated by exposure to an ongoing sequence of states, and monitoring the subject's physiological data. Method. The physiological data comprises waveform data and estimating the CRI;
Comparing the waveform data with a plurality of sample waveforms, wherein each of the sample waveforms generates a similarity coefficient that represents a similarity between the waveform data and each of the sample waveforms; Steps corresponding to different values;
Normalizing the similarity factor for each of the sample waveforms;
Summing the normalized similarity coefficients to generate an estimated CRI value for the patient;
The method of claim 4 comprising:   The method of claim 2, wherein estimating the probability that the patient is bleeding is based on a fixed time history that monitors the physiological data of the patient.   The method of claim 2, wherein estimating the probability that the patient is bleeding is based on a dynamic time history monitoring the physiological data of the patient.   At least one of the one or more sensors is a blood pressure sensor, an intracranial pressure monitor, a central venous pressure monitoring catheter, an arterial catheter, an electroencephalograph, a heart monitor, a transcranial Doppler sensor, a transthoracic impedance plethysmograph, a pulse oximeter, 3. The method of claim 2, wherein the method is selected from the group consisting of a near infrared spectrometer, a ventilator, an accelerometer, an electrical eye diagram, a transcutaneous glucometer, an electrolyte sensor, and an electronic stethoscope.   The method of claim 2, wherein the physiological data comprises blood pressure waveform data.   The method of claim 2, wherein the physiological data comprises plethysmographic waveform data.   The method of claim 2, wherein the physiological data comprises photoplethysmograph (PPG) waveform data.   The method of claim 2, wherein analyzing the physiological data comprises analyzing the physiological data against an existing model.   24. The method of claim 23, further comprising generating the existing model prior to analyzing the physiological data. Generating the existing model comprises:
Receiving data relating to one or more physiological parameters of the subject to obtain a plurality of physiological data sets;
Using a reference sensor to directly measure one or more physiological states of the subject to obtain a plurality of physiological state measurements;
Correlating the received data with the physiological state measurement of the subject;
25. The method of claim 24, comprising:   26. The method of claim 25, wherein the one or more physiological conditions comprises a decrease in circulatory system volume.   27. The method of claim 26, further comprising inducing the physiological condition of reduced circulatory volume in the subject.   28. The method of claim 27, wherein inducing the physiological condition comprises exposing the subject to lower body negative pressure (LBNP).   29. The method of claim 28, wherein inducing the physiological condition comprises subjecting the subject to dehydration.   26. The method of claim 25, wherein the one or more physiological conditions comprise a condition close to or close to cardiovascular collapse.   26. The method of claim 25, wherein the one or more physiological conditions comprise a normal fluid volume condition.   26. The method of claim 25, wherein the one or more physiological conditions comprise a condition of excessive circulating blood volume.   26. The method of claim 25, wherein the one or more physiological conditions comprise a dehydration condition. Correlating the received data with the physiological condition measurement of the subject;
For each of one or more results _{o k,} signals _{s 1, s} 2,. . . Comprising the steps of identifying the most predictive set of signals S _k of the set of s _D, the most predictive set of signal S _k is a first data set representing a first physiological parameter corresponds to each of the one or more results o _k is representative of the physiological condition measurements, the steps,
Stochastic prediction model
Autonomously learning a set of
There is a prediction of the result o _k resulting from the model M _k to be used as an input value obtained from the set of signals S _k, the steps,
Signals s ₁ , s ₂ ,. . . , S _D and the corresponding results o ₁ , o ₂ ,. . . Repeating the step of learning autonomously gradually from data including examples of values of o _K ;
26. The method of claim 25, comprising: A device,
A non-transitory computer readable medium having encoded thereon a set of instructions executable by one or more computers, the set of instructions on the device;
Receiving physiological data from one or more sensors;
Analyzing the physiological data;
Estimating the probability that the patient is bleeding,
Displaying at least one of an evaluation, a predicted value, or an estimated value indicating a probability that the patient is bleeding on a display device;
apparatus.