SU843941A1 - Device for intensive supervision of seriously ill patients - Google Patents

Description

tions. Display devices; a visual information display unit; a binary code converter to a binary-decimal printing unit. Auxiliary devices include a communication interface, timers, a clock generator, an instruction block. In addition, the device contains a command generation unit and a signal-spike unit (2.

During operation of the known device in the continuous tracking mode, dynamic increments of the patient's state vector into the storage unit are made. The state vector is the combination of the parameters that describe the state of the organism at the current time. Each of the parameters can take one of several values, for example, from O to 9, in accordance with the coding table, an example of which is given in Table. I. Vectors are divided into classes. For the case of resuscitation patients, the first clashes form vectors of the state of surviving patients. The choice of the treatment factor is made by the operator's instructions on the basis of the cell information in the storage device of the dynamic increments of the state vector. However, during the operation of the device, a sharp deterioration of the patient's condition may occur, which is reflected in a change in the state vector, but the known device does not respond to this situation.

Sharp deterioration of the patient's condition, like Iravilo, is preceded by changes in individual physiological parameters to the limits that are not yet caused by the sigtalus. However, when evaluating changes in the state vector, it is possible to predict the outcome of the disease for various periods. The prognosis improves patient care, warns the doctor about a possible worsening of the condition of the patient. The second application of predicted changes in the patient's state vector is associated with the problem of generating a statistically reliable alarm using a set of parameters. Criteria for issuing an alarm signal. H.S. MUST be dynamically adjusted.

. It must be borne in mind that the variation of individual dimensions also has an irogic value. For example, some signs of an ECG signal, obtained as a result of statistical processing, can predict the development of deadly complications (acute conditions). On the other hand, the results of processing the cardio signal are one of the complex features that is part of the state vector.

Thus, the known device does not provide prediction of the patient’s condition.

and generate a statistically valid alarm.

The aim of the invention is to ensure the formation of a timely therapeutic effect on the patient by making a forecast based on the results of his analysis, as well as forming a statistical one. reliable at the moment alarm signal.

The goal is achieved by the fact that a device containing bioinformation sensors, an amplifier, an analog switch from a switch, an analog-digital converter, a switch on discrete information, a generator of a constant vector with a block of constant data, a total offset unit, a block of information representation, and a converter binary code in binary decimal and

digital alarm unit, sequentially connecting the block of selecting the optimal treatment factor, whose outputs are combined with a communication interface, a set of permissible therapeutic effects and an anti-imbalance mechanism, a timer, a clock generator and a block of operator commands that are connected to the inputs of the command generation unit, as well as a block signaling, the input of the block of selection of the optimal therapeutic factor is connected to the output of the total memory block, and the output of the block of the set of permissible therapeutic effects with the second input an actuator, an electrocardiogram amnlitud analyzer and a prognostic sigpal allocation unit, interconnection sequence, information analysis block and dynamic correction block were introduced, and the ip formation analysis block consisted of medical memory blocks, threshold schelta, response unit block, unified communication unit, and the unit and its own and group. and the diyamic adjustment block consists of a selection block

state vectors, locator of the nerve klaes vector block and second vain block of second class vectors, decisive block and information rewriting block, connected in series, with the first inputs of medical memory blocks in the information analysis block connected to the output of the state vector formator, the second inputs of medical blocks and these are connected to the corresponding outputs of the command generation unit, and the third inputs of the medical gamut units are connected to the output of the rewriting unit in the dynamic adjustment unit, you the course of the reaction analysis block is connected through the block of selection of state vectors to the inputs of the block of the auto financing of the first class vectors and the block of storing the vectors of the second class, the outputs of which are connected to the corresponding inputs of the first

block, and the second input of the state vector block is connected to the output of the total memory block, the output of the alarm unit is connected to the second input of the optimal therapeutic factor selection block, the first input of the alarm block is connected to the output of the threshold circuit, and the second input to the output of the prediction signals selection block and To the input of the shaper of the state vector, the input of the amplitude analyzer of the electrocardiogram is connected to the output of the switch of discrete information. The drawing shows a block diagram of a device for intensive monitoring of diseased patients. The device for intensive monitoring of T-bar contains a block of 1 sensors, sensors 2 of bionic information, amplifier 3, switch 4 analog signals, analog-to-digital converter 5, switch 6 of discrete information, analyzer 7 amplitudes of prognostic signals, conditioner 9 , block 10 of visual information representation, binary code converter 11 into twofold-ten-digit, digital-printing block 12, block 13 for choosing the optimal therapeutic factor, communication interface 14, block 15 sets of permissible therapeutic effects, actuator 16, timer 17, command generation unit 18, clock generator 19, operator command unit 20, medical memory unit 21, threshold circuit 22, alarm unit 23, reaction analysis unit 24, unit 25 selection of state vectors, storage unit 26 nepisoro class vectors, storage unit 27 second class vectors, decision unit 28, information rewriting unit 29, fixed data unit 30, total storage unit 31, blocks 21, 23, 24 and circuit 22 information analysis block, and ki 25, 26, 27, 28, 29 are combined in block dynamic corrections medical memory. Block 1 sensors are connected in series with sensors 2. amplifier 3, switch 4 analog signals, analog-digital converter 5, switch 6 discrete information. The latter is connected to the shaper 9 of the state vector, the amplitude analyzer 7 and the block 10 of visual information representation. The data block 30 is connected to the state vector shaper 9, which is sequentially connected to the information analysis unit, the medical memory dynamic correction unit and the total memory unit 31. Specifically, the state vector shaper 9 is connected to the medical memory blocks 21, which are connected with a threshold circuit 22 connected to the alarm unit 23 and the reaction analysis unit 24. The signaling unit 23 is connected with the optimal therapeutic factor selection unit 13, the outputs of which are connected to the set 15 of the allowable therapeutic effects, the actuator 16n by the communication interface 14. The state vector selection unit 25 is connected to the total storage unit 31, the reaction analysis unit 24 and the storage units 26, 27 of the first and second class vectors, which are connected to the deciding unit 28 and the information rewriting unit 29 connected to the medical memory blocks 21. The amplitude analyzer 7 is connected to a prediction signal extraction unit 8, which is connected to the alarm unit 23 in the condition vector shaper 9. A visual information presentation unit 10, a binary code to binary-decimal converter 11 and a digital-printing unit 12 are connected in series. A timer 17, a command generation unit 18, a clock pulse generator 19 and an operator command unit 20 are auxiliary, with the command generation unit 18 being connected to the medical memory units 21. The device works as follows. Medical memory blocks contain information in the form of prognostic matrix coefficients and linear regression equation coefficients. The prognostic matrix A (a; y) contains p (1) coefficients a ,; - In - | -, where i is the sign number, j is the sign gradation, P. .. is the frequency of occurrence of the j-th value of the i-Toro sign in the first class, is the frequency of the j-th i-Toro value of the sign in the second classroom. In the simplest case, the first class is formed by the vectors of the state of the deceased patients, and the second class is formed by the vectors of the state of the surviving patients. The coefficients aij are formed on the basis of processing arrays of verified case histories selected for a given disease, traumatists nanrnlter, characterized by moments of the disease, for example, for traumatizers at the time of first aid, after the first day, after the first day, etc. According to these solutions, forming the | nm state vector (xi, ..., x ,,), the sum of the coefficients Sv a, -y is calculated. According to the value of S, the trial is based on the outcome of the disease: if 1, then the outcome is unfavorable, if S M 2, then the outcome is favorable, if Vll S M 2, then the decision on the prediction of the outcome is not taken. M 1 and M 2 are thresholds that are a function of a given prediction accuracy depending on the amount of clinical material. The thresholds M 1 and M 2 are pre-calculated - and refined experimentally. The sum S is calculated in the threshold device 22, with the first and the third outcome giving an alarm to the alarm unit 23, or not having such a CHJ-cash, and in the reaction analyzing unit 24 confirming or not confirming the correctness of the alarm. Due to the fact that the device is intended for use in resuscitation-anesthesiology and intensive care units, the choice of features for the prognostic matrix is focused on those features that allow you to detect a sudden cardiac arrest or warn of such a threat. Such signs are; pulse rate, respiration rate, body temperature, systolic and diastolic blood pressure, ECG signal. The range of change of a trait can be divided into a number of sub-bands. An example of the coding table, in which the distinction between the signs for the prognostic matrix is given, is given in Table. 1. Based on computer processing of case histories, prognostic matrices A were obtained for traumatic patients at the time of first aid. An example of such a prognostic

Tables of the matrix containing the coefficients of the matrix A - {Yau) are given in Table 2 for the forecast for 5 hours. 2. A second type of medical memory which is recorded in blocks 21 ,. can be represented in the form of linear coefficients of the VU. a, x / + ao,; -1 multiple regression coefficients; X; - a sign that can take on some meaning. Other types of coefficients obtained as a result of differences in the types of mathematical processing of clinical material can be recorded in blocks 21 of medical memory. For example, coefficients of a multifactorial dispersion complex or conditional probabilities of disease with this symptom can be used as a medical memory. All kinds of medical memories are pre-recorded in blocks 21. In threshold scheme 22, thresholds M I and M 2 are stored for each medical memory unit 21, the sum of 2 a is calculated and compared with the corresponding thresholds.

Sign of

2 3 4 5 6

Claims (2)

Table 2 The state vector is transmitted to the medical memory blocks periodically. The period is set by the operator and the signal from timer 17. The same signals are used to record the state vector in the total memory block 31. The amplitude analyzer of the ECG signal allows to obtain the distribution of the biopotentials of the heart, which, as is well known, makes it possible to register all those changes that on the ECG, they are expressed by a deformation of the normal curve: a change in the polarity of the teeth, a broadening of the ORS complex, a deepening of the negative teeth. Such numerical characteristics of the distribution function, which are present at the output of the analyzer 7 as a mode, its numerical value, location and the presence of the second mode, have a prognostic value. The prediction signal extraction unit 8 controls the numerical characteristics. An alarm is issued by block 23 in the following cases if: the numerical value of the main mode is more than 0.7 or less than 0.5 of the total amplitude values analyzed; the position of the main mode is shifted towards the higher bits of the histogram, starting with a half value (for example, 5, 6, 7 ... 12 bits with 12 classes of the histogram); the presence of the second mode in any class, if it reaches 3/4 of the magnitude of the main mode. In addition, the magnitude of the ECG signal is monitored. Block 8 provides an alarm if the magnitude of the ECG signal is less than 0.3 mV. The listed numerical characteristics of the distribution function of the biopotentials of the heart are an integral part of the patient's condition vector. The dynamic correction module of the medical memory provides for the correction of the contents of the medical memory blocks 21 as the clinical information accumulates in the total memory block 31. In accordance with the content of the medical radiation blocks from the total storage unit 31 samples the state vectors under control of the state vector selection unit 25, and roizvodits type of disease, in specific moments disease by deadline prognosis. After recording the state vectors belonging to the first and second classes, the storage units 26 and 27 are used to calculate the coefficients of the prediction matrices Zg / In in the decision block 28, which are recorded using the local information rewriting lock 29 into the medical memory blocks 21. Physiological information about the state of the patient from the sensor block 1, removed using sensors 2, after amplification and normalization in amplifier 3 and passing through switch 4, enters the input of the digital-to-digital converter 5 and further to switch 6 of discrete information, from the output of which information arrives a 7 amplitude analyzer, a visual information display unit 10, and a state vector generator 9. The permanent data from block 30 is fed to another input of the former of the 9th state vector. In the intensive monitoring mode, the state vectors of the generator 9 are fed to the input of the total memory block 31, as well as to the inputs of the medical memory blocks 21, the outputs of which are connected to a threshold circuit 22, in which the prediction sum is calculated and compared with the thresholds M 1 and M 2. if the predictive amount goes beyond one of the thresholds M 1 or M 2, as well as in case of failure to make a decision, the alarm unit 23 is triggered, which starts the block 13 for selecting the optimal treatment factor, block 15 of the set of therapeutic factors The body mechanism 16, with the help of which the therapeutic infusion is produced. A call to the reaction analysis block is made directly from the threshold block. The selection of the medical memory block or their sequence is made by commands from the command generation unit 18. The output of the reaction analysis unit 24 is at the same time the output of the information analysis unit connected via the input of the state vector selection unit 25 with the dynamic adjustment unit. In the dynamic correction mode, medical memory from the total storage unit 31 in block 25 selects the state vectors belonging to the first and second classes according to the commands from the reaction analysis unit 24. The vectors are recorded in the storage blocks 26, 27 of the vectors of the first or second classes, are transferred to the decision block 28, where the prediction matrix elements are calculated and then to the information rewriting block 29, which provides for the correction of information in the medical memory blocks 21. Thus, on the one hand, information from block 31 goes simultaneously to block 10 and block .13 of the choice of the optimal treatment factor, and on the other hand is subjected to additional processing in the block of dynamic adjustment, the ECG signal from the output of switch 6 of discrete information enters the analyzer 7 amplitudes , the output of which is connected to the prediction signal extraction unit 8, which, when using a predictive prediction, includes an alarm unit 23 and a therapeutic circuit including blocks 13, 14, 15 and an actuator 16. In the absence of signs of an unfavorable prediction, the numerical characteristics of the amplitude distribution of the ECG signal are a component of the state vector in the conditioner 9, the state vector. The control commands generated in the command generation unit 18 are synchronized with the temporal characteristics of the applied influences determined by the timer set 17. The communication interface 14 provides for the exchange of information with the control computer. Timer 17, blocks 18, 20, and generator 19 provide polling and generation of state vectors at predetermined intervals (ECG, pulse rate, etc.). The visual information display unit 10 allows the operator to control the operation of the device. In the binary-to-binary converter code 11, docking is carried out with the digital-printing unit 12, which records information about the treatment process. As a result of the experiments, it was found that for resuscitation patients the most rational forecasting is at 1 and 5 hours, but the real device allows prediction at 24, 120 and 240 hours. The state vector is formed according to a different number of signs: 39, 17, 15, 10 , 6 and 5, depending on the class of the disease and the condition of the patient. The period of formation of the state vectors and the evaluation of the ECG are performed once every 2 minutes or at other times as directed by the physician. The forecast period and its type depending on the disease is also prescribed by the doctor. Dynamic adjustment of medical memory is made as the clinical material is accumulated, at least once a week. Alarms are issued to signaling units located in rooms for medical personnel on duty. A number of device blocks: a total memory block, a decision block, medical memory blocks, and some others can be made on the basis of an Minsk-32 type computer. Thus, since the device makes a dynamic adjustment of the forecast criterion and alarm criterion, as well as the processing of the ECG signal, designed to highlight the signs that have an independent prognostic value and are part of the vector of the patient's state, the device provides a forecast based on the results of the analysis of the patient's state vector with the issuance of a statistically reliable alarm at the moment. As a result, the quality of patient care improves, and the work of the physician is facilitated, since a number of monotonous, repetitive functions for evaluating information that comes from the patient is transmitted to a device that generates alarm signals for an unpleasant prognosis. In addition, the efficiency of patient care is increased. The contents of the medical memory blocks are dynamically adjusted to ensure the required quality of the device operation. Claims of the Interstate Public Device for Target Diseases, containing bioinformation sensors, s silica gel, analog signal switcher, analog-to-digital converter, discrete information switch, constant-state data generator with a block of constant data, total memory block, visual information display unit, binary converter code in binary-decimal and digital-printing unit, connected in series, the block of choice of the optimal therapeutic factor, the outputs of which are connected to a communication interface, a set of admissible therapeutic effects and an actuator, a timer, a clock pulse generator and an operator command block connected to the inputs of the command generation unit, and also an alarm unit, the input of the optimal therapeutic factor selection block connected to the output of the total memory block and the output of the set of permissible therapeutic effects with the second input of the actuator, characterized in that, in order to form a timely therapeutic effect on the patient by processing the forecast based on the results of its analysis, as well as generating a statistically reliable alarm at the moment, an electrocardnogram analyzer and a prediction signal extraction unit are inserted into the device, connected in series to an information analysis unit and a dynamic adjustment unit, the information analysis unit consisting of medical units iam tee, threshold scheme, reaction analysis block, connected in series, and the dynamic adjustment block consists of the vector selection block s, the storage unit vectors of the first class and zapomi} 1ayuschego unit vectors Class II reschayuschego unit and rewriting unit iiformatsii connected in series. the first inputs of the medical memory blocks in the information analysis block are connected to the output of the state vector shaper, the second inputs of the medical memory blocks are connected to the corresponding outputs of the command generation unit, and the third inputs of the medical memory blocks are connected to the output of the rewriter block from the dynamic adjustment block, the output of the reaction analysis unit is connected via the state vector selection unit to the inputs of the first class vector storage unit and the second class vector storage unit, the outputs of which are connected Ina with corresponding 1, the inputs of the decision block, and the second input of the state vector selection block is connected to the output of the total memory block, and the output of the alarm unit is connected to the second input of the optimal therapeutic factor selection block, the first input of the signal block is connected to the output of the threshold circuit, and the second an input to the output of the prognostic signals selection unit and to the input of the state vector shaper, and an analogue input of the amplitudes of the electrocardiogram connected to the output of the switch of discrete information. Sources of information taken into account in the examination 1. A hierarchical system of intensive observation based on mini-computers. Express information IN THEM. Seri "Mednitsiska technique, 1977, No. 7, p. 17-27. 2. Frolov A. B. and others. Automated control of treatment. “Medical equipment, 1977, No. 4, p. 5-9 (prototype).