CN118173272B - Method for determining risk level and carrying out early warning through attenuation of SOFA score - Google Patents
Method for determining risk level and carrying out early warning through attenuation of SOFA score Download PDFInfo
- Publication number
- CN118173272B CN118173272B CN202410591821.0A CN202410591821A CN118173272B CN 118173272 B CN118173272 B CN 118173272B CN 202410591821 A CN202410591821 A CN 202410591821A CN 118173272 B CN118173272 B CN 118173272B
- Authority
- CN
- China
- Prior art keywords
- ward
- image
- round
- information
- voice
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000012544 monitoring process Methods 0.000 claims abstract description 162
- 238000007689 inspection Methods 0.000 claims abstract description 117
- 238000012360 testing method Methods 0.000 claims abstract description 41
- 238000006243 chemical reaction Methods 0.000 claims abstract description 25
- 238000012545 processing Methods 0.000 claims abstract description 11
- 238000000605 extraction Methods 0.000 claims abstract description 5
- 230000008859 change Effects 0.000 claims description 102
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000011156 evaluation Methods 0.000 claims description 8
- 238000011161 development Methods 0.000 claims description 4
- 230000003321 amplification Effects 0.000 claims description 3
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 3
- 238000012806 monitoring device Methods 0.000 claims description 2
- 210000000056 organ Anatomy 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 210000001835 viscera Anatomy 0.000 description 3
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 210000001124 body fluid Anatomy 0.000 description 2
- 239000010839 body fluid Substances 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 208000028399 Critical Illness Diseases 0.000 description 1
- 206010053159 Organ failure Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 235000016709 nutrition Nutrition 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/74—Details of notification to user or communication with user or patient ; user input means
- A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- Biomedical Technology (AREA)
- Primary Health Care (AREA)
- Molecular Biology (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- Physiology (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- Heart & Thoracic Surgery (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a method and a system for determining attenuation risk level and performing early warning through SOFA scoring. The method comprises the following steps: step S1: acquiring a monitoring image of a patient in real time, acquiring first ward-round voice in a preset time period, and acquiring a checking image in time; step S2: converting the second ward-round voice into ward-round information through a semantic conversion unit; step S3: extracting test data from the test image by an extraction unit; step S4: comparing the monitoring image with a reference monitoring image according to ward-round information and inspection data to obtain a first image closest to the monitoring image, and obtaining monitoring information of the first image; step S5: and scoring according to the monitoring information, the ward-round information and the inspection data to determine the risk level and carrying out early warning based on the risk level. The invention solves the problem of inaccurate SOFA scoring and improves the SOFA scoring accuracy.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a system for determining attenuation risk level and performing early warning through SOFA scoring.
Background
With the progress of technology, the medical technology and the level are continuously improved, and the critical disease quality level and ICU construction of China are greatly developed and improved. In current clinical treatment efforts, the severity score is closely related to the severity of the patient's condition. Through reasonable and scientific severe scoring, the improvement, development and treatment effect of the patient can be comprehensively and comprehensively evaluated. An accurate severe scoring result is of great significance and value for clinical treatment. For example: chinese patent CN109524124a, the invention discloses a severe scoring system, comprising: an acquisition module for acquiring data values of parameters of each scoring item; an automatic calculation scoring module for automatically calculating a score based on the worst value of the acquisition parameters; a variety of severe medical scores may be provided, including ApacheII, SOFA, VTE, nutritional scores. Taking ApacheII scoring as an example, the method can be used for automatically selecting the worst value corresponding to each parameter to calculate the score in a selected time range; a graphic display module for displaying to the user a score range of each parameter value variation trend and the score value of the score item within the selected time period; a GCS score recording module for inputting GCS score records by a user. The method adopts a chart combining mode, displays the acquired data and the scoring standard in a visual chart display mode, and can automatically calculate the score or manually modify and calculate the score of the data in the selected time range, thereby reducing the scoring fault tolerance rate and effectively improving the working efficiency of clinical workers. Also for example: european patent WO2018073646A1, a system for predicting Sequential Organ Failure Assessment (SOFA) scores using machine learning, which may process features using one or more SOFA score prediction models derived from at least one machine learning process to output corresponding predicted SOFA scores. One predictive model has been trained to output a first SOFA component score over a first time period in the future and a second predictive model has been trained to output a second SOFA component score over the first time period in the future. The system may output a total SOFA score, a first SOFA component score, on a graphical user interface. Both of the above patents determine the risk level of a critically ill patient by SOFA scoring, but do not take into account the comprehensiveness and accuracy of the patient scoring data, and therefore do not have high accuracy and cannot score in real time to obtain the risk level.
Disclosure of Invention
In order to better solve the above problems, the present invention provides a method for determining attenuation risk level and performing early warning through SOFA scoring, the method includes the following steps:
step S1: acquiring a monitoring image of a patient in real time through an acquisition unit, acquiring first ward round voice in a preset time period, acquiring a test image in time, and preprocessing the monitoring image and the test image;
Step S2: denoising the first ward-round voice to obtain second ward-round voice, obtaining first text information from the second ward-round voice through a semantic conversion unit, extracting patient information from the first text information, continuously receiving third ward-round voice, obtaining fourth ward-round voice through denoising, converting the fourth ward-round voice into second text information through the semantic conversion unit, obtaining ward-round information from the second text information, storing the third ward-round voice, and storing the third ward-round voice;
step S3: extracting test data from the test image by the extraction unit;
Step S4: according to the ward-round information and the inspection data, comparing the first image closest to the monitoring image based on the monitoring image and a reference monitoring image, and acquiring monitoring information of the first image;
Step S5: and scoring according to the monitoring information, the ward-round information and the inspection data to determine a risk level and carrying out early warning based on the risk level.
As a preferable technical scheme of the invention, the monitoring image is a display image for monitoring and obtaining a monitoring result through monitoring equipment, the inspection image is an inspection result image, and the first ward-round voice is an evaluation dialogue for the disease development of a patient when a doctor performs ward-round.
As a preferable embodiment of the present invention, in the step S1, the monitoring image and the inspection image are preprocessed: performing angle adjustment and binarization on the first monitoring image; the inspection image is enhanced.
As a preferred technical solution of the present invention, step S2 includes the steps of:
Step S21: periodically acquiring the first ward-round voice in a preset time period through the acquisition unit, taking the last acquired first ward-round voice as first background noise, and preprocessing the first ward-round voice based on the first background noise to acquire the second ward-round voice with the first background noise removed;
Step S22: converting the second ward-round voice into first text information through a semantic conversion module, analyzing the text information into words, and storing the second ward-round voice when the words of the first text information comprise identification information of a patient;
Step S23: the first ward-round voice is continuously and periodically obtained to serve as a third ward-round voice, the third ward-round voice is sliced to obtain a first slice voice, the first slice voice is compared with the second ward-round voice, the slice voice which does not contain the voiceprint corresponding to the second ward-round voice is taken as second background noise, and the second background noise is removed from the slice voice which is closest in subsequent distance time and contains the voiceprint corresponding to the second ward-round voice to obtain a second slice voice;
Step S24: and splicing the second sliced voice to obtain fourth ward-round voice, converting the fourth ward-round voice into second text information through the semantic conversion unit, and extracting the ward-round information from words analyzed by the second text information.
As a preferred embodiment of the present invention, the step S3 includes the following steps:
Step S31: acquiring a test image template matched with the test image based on the distribution of tables, characters, numbers and figures in the test image and the area occupied by the test image;
Step S32: dividing the inspection image into a plurality of sub-inspection images based on the position information of each inspection object in the inspection image template, comparing each sub-inspection image with each reference image, and obtaining a plurality of target images;
Step S33: and converting a plurality of the target images into a test text by a conversion unit, and extracting the test data from the test text.
As a preferred embodiment of the present invention, the step S4 includes:
Comparing the test data with recent historical test data, judging a first risk change of the patient according to a comparison result, comparing the first risk change with a second risk change obtained by comparing last ward round information with current ward round information, and obtaining a comparison result, when the comparison result is consistent with the first risk change and the second risk change, increasing the identification precision of the monitoring image to be first precision when the first risk change and the second risk change are attenuation risks and rise, and decreasing the identification precision of the monitoring image to be second precision when the first risk change and the second risk change are attenuation risks and fall, and when the first risk change and the second risk change are attenuation risks and invariable, the identification precision of the monitoring image is invariable, wherein the first precision is larger than the second precision;
And carrying out set proportion amplification on the monitoring image and the reference monitoring image, wherein the higher the identification precision of the monitoring image is, the larger the corresponding set proportion is, the grid division is carried out on the monitoring image and the reference monitoring image, the grid comparison is carried out on the monitoring image and the reference monitoring image one by one, the first image closest to the monitoring image in the reference monitoring image is obtained, and the monitoring information corresponding to the first image is obtained.
As a preferred embodiment of the present invention, the step S4 further includes: and when the first risk change is inconsistent with the second risk change, acquiring a first average value and a second average value of the monitoring information of the previous day and today, comparing the first average value with the second average value, acquiring a third risk change, when the second risk change is inconsistent with the third risk change, acquiring the ward-round information again through the stored third ward-round voice and the method of the step S2, when the first risk change is inconsistent with the third risk change, acquiring the inspection data again through the step S3, and setting the recognition precision of the monitoring image according to the third risk change.
As a preferred embodiment of the present invention, the step S5 includes: and scoring the monitoring information, the ward-round information and the inspection data through a SOFA scoring rule, determining the risk level based on the scoring, and carrying out early warning according to the risk level.
As a preferred embodiment of the present invention, the first risk change and the second risk change are both risk changes of attenuation of the patient on the same day relative to attenuation of the patient on the previous day.
As a preferred embodiment of the present invention, the ward-round information includes SOFA scoring item information other than the inspection data and the monitoring information.
The invention also provides a system for determining attenuation risk level and carrying out early warning through SOFA scoring, wherein the system is used for realizing the method, and the system comprises the following steps:
the acquisition unit is used for acquiring a monitoring image of a user in real time, acquiring a checking image and first ward-round voice in time, and preprocessing the monitoring image and the checking image;
the processing unit is configured to: denoising the first ward-round voice to obtain a second ward-round voice, obtaining first text information by the second ward-round voice through a semantic conversion unit, extracting patient information from the first text information, continuously receiving a third ward-round voice, and obtaining a fourth ward-round voice through denoising; the method comprises the steps of comparing the monitoring image with a reference monitoring image according to ward-round information and inspection data to obtain a first image closest to the monitoring image, and obtaining monitoring information of the first image; extracting inspection data from the inspection image;
the semantic conversion unit is used for converting the fourth ward-round voice into second text information, acquiring ward-round information from the second text information and storing the third ward-round voice;
And the early warning unit is used for scoring according to the monitoring information, the ward-round information and the inspection data to determine the risk level and carrying out early warning based on the risk level.
Compared with the prior art, the invention has the following beneficial effects:
The invention acquires the first ward round voice in a ward round preset time period, acquires the second ward round voice through the first ward round voice denoising processing, acquires first risk change through converting the second ward round voice into first text information, further acquires third ward round voice, slices the third ward round voice, identifies second background noise, denoising processing the first slice voice, acquires second slice voice, splices the second slice voice to acquire fourth ward round voice, acquires the fourth ward round voice through a semantic conversion unit, acquires ward round information through the second text information, acquires inspection data through the inspection image, acquires first risk change through comparing inspection data of a day before a patient with inspection data of the day before the patient, acquires second risk change through comparing the ward round information of the day before the patient with the ward round information, acquires second risk change based on the first risk change and the second risk change, acquires the image attenuation information according to the recognition plan, and can further reduce the attenuation of the inspection image according to the first risk change, and the image attenuation of the inspection image can be correspondingly reduced when the image is detected, and the image attenuation is detected by the image attenuation information of the patient is detected according to the first detection image.
Drawings
FIG. 1 is a flow chart of a method for determining attenuation risk level and performing early warning according to SOFA score;
FIG. 2 is a block diagram of a system for determining attenuation risk level and performing early warning according to SOFA score.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The invention provides a method for determining attenuation risk level and carrying out early warning through SOFA scoring, which comprises the following steps as shown in figure 1:
step S1: the monitoring image of the patient is acquired in real time through the acquisition unit, the first ward-round voice is acquired in a preset time period,
Acquiring a test image in time, and preprocessing the monitoring image and the test image;
Specifically, the acquisition unit acquires, in real time, a monitoring image of the user, the monitoring image being from a monitoring device for monitoring vital signs of a patient, for example: respiration monitoring equipment, heart brain monitoring equipment and the like, wherein the first ward round voice is from the evaluation of the current situation of the patient by the corresponding attending doctor of the patient in the daily ward round time. The inspection image is an inspection result obtained by testing blood or other body fluids, the monitoring image, the inspection image and the first ward-round voice are processed to obtain the monitoring information, the ward-round information and the inspection information of the patient, and the attenuation risk level of the user is accurately obtained in real time through scoring and early warning is carried out timely.
Step S2: denoising the first ward-round voice to obtain second ward-round voice, obtaining first text information from the second ward-round voice through a semantic conversion unit, extracting patient information from the first text information, continuously receiving third ward-round voice, obtaining fourth ward-round voice through denoising, converting the fourth ward-round voice into second text information through the semantic conversion unit, and obtaining ward-round information from the second text information;
Specifically, because there are multiple patients in an ICU ward and there are various kinds of noise emitted by instruments, when the attending doctor looks at the patient and evaluates the patient, the evaluation information of the patient cannot be accurately acquired, so that the background noise of the first ward voice is particularly critical, therefore, in a preset time period corresponding to the ward, the first ward voice acquired by the next time is periodically acquired, and the first ward voice acquired last time is taken as the first background noise, the first ward voice acquired this time is taken as the second background noise, the first ward voice is removed from the second ward voice, the first text information is acquired from the second ward voice through semantic conversion, the first text information is parsed into words, and when the words include the identification information of the patient, the identification information can be the name of the patient or the position number of the patient, the first ward voice acquired next time is determined to correspond to the patient, so that the second ward voice acquired by the first ward voice is saved, the second ward voice is accurately acquired as the second background noise, the second text information is acquired from the second ward voice is accurately compared with the second text information, the second text information is acquired by the second text information, the second text information is processed by slicing, the second text information is acquired from the second ward voice is accurately, and the second text information is acquired, the second text information is processed, and the second text information is acquired is cut into the second text information, and lay a foundation for accurately acquiring the identification precision of the monitoring image and the risk level of the patient.
Step S3: extracting test data from the test image by the extraction unit;
Specifically, since different inspection images correspond to different formats, for example: the method comprises the steps that the reference object of the inspection image obtains an inspection image template matched with the inspection image from the storage unit, wherein the reference object can be in a table format and text distribution in the inspection image, position information corresponding to each inspection item is obtained based on the inspection image template, the inspection image is divided into a plurality of target images based on the position information, the target images are converted into inspection texts, the target images are inspection items corresponding to SOFA scores, inspection data are extracted from the inspection texts, and accurate inspection data can be obtained and basis is provided for obtaining the identification precision of the monitoring image and the risk level of a patient through the technical scheme.
Step S4: according to the ward-round information and the inspection data, comparing the first image closest to the monitoring image based on the monitoring image and a reference monitoring image, and acquiring monitoring information of the first image;
specifically, the organ decay risk change of the patient can be obtained through the monitoring information and the ward information, when the risk changes reflected by the monitoring information and the ward information are both attenuation risk increases, the risk of the patient is larger, and because vital signs of the patient pass through, the identification precision of the monitoring image is increased, more accurate monitoring information is obtained so as to inform medical staff to take corresponding measures in time.
Step S5: and scoring according to the monitoring information, the ward-round information and the inspection data to determine a risk level and carrying out early warning based on the risk level.
Specifically, all the SOFA scoring items can be obtained through the monitoring information, the ward-round information and the inspection data, and the real-time monitoring information can be further obtained through the monitoring images obtained in real time, so that the accurate SOFA scoring can be obtained through the technical scheme, the SOFA scoring can be obtained in real time, the risk level is determined based on the scoring, early warning is carried out based on the risk level, corresponding measures are timely taken, and deterioration of visceral organs of a patient is prevented.
Further, the monitoring image is a display image of a monitoring result obtained through monitoring by monitoring equipment, the inspection image is an inspection result image, and the first ward-round voice is an evaluation dialogue for the disease development of the patient when a doctor performs ward-round.
Specifically, the comprehensive scoring reference information can be obtained through the inspection data obtained through the inspection image, the monitoring information obtained through the monitoring image and the ward-round information obtained through the first ward-round voice, and a data basis is provided for obtaining the accurate attenuation risk level of the patient.
Further, in the step S1, the monitoring image and the inspection image are preprocessed: performing angle adjustment and binarization on the first monitoring image; the inspection image is enhanced.
Specifically, the angle of the monitoring image is consistent with the angle of the reference monitoring image by preprocessing and adjusting the angle of the monitoring image due to the problem that the angle of the monitoring image is different from the angle of the reference monitoring image, so that subsequent recognition processing is facilitated, the inspection image is enhanced, and the problem that the inspection image is not matched with a proper inspection image template due to unclear display is avoided.
Further, step S2 includes the steps of:
Step S21: periodically acquiring the first ward-round voice in a preset time period through the acquisition unit, taking the last acquired first ward-round voice as first background noise, and preprocessing the first ward-round voice based on the first background noise to acquire the second ward-round voice with the first background noise removed;
Specifically, since there are multiple patients in an ICU ward and there are various kinds of noise emitted by instruments, when the attending doctor checks the patient and evaluates the attenuation condition of the organs of the patient, the evaluation information of the patient cannot be accurately obtained, so that the background noise of the first ward voice is particularly critical, therefore, in the preset time period corresponding to the ward, the first ward voice is periodically obtained, the last obtained first ward voice is taken as the first background noise, the first background noise is removed from the first ward voice, the second ward voice is obtained, whether the words in the first text information corresponding to the second ward voice include the identification information of the patient or not is confirmed, and whether the following first ward voice is the attenuation risk evaluation of the patient or not is confirmed.
Step S22: converting the second ward-round voice into first text information through a semantic conversion module, analyzing the text information into words, and storing the second ward-round voice when the words of the first text information comprise identification information of a patient;
Specifically, when the attending physician performs the attenuation risk assessment on the patient, since one ICU ward may hold a plurality of patients, the patient information may be checked first, by using the above technical solution, not only the first ward-round voice corresponding to the patient, that is, the third ward-round voice obtained after the first ward-round voice corresponding to the first text information including the patient identification information, may be obtained, but also the second ward-round voice may be stored, so as to provide a reference for obtaining the second background noise.
Step S23: the first ward-round voice is continuously and periodically obtained to serve as a third ward-round voice, the third ward-round voice is sliced to obtain a first slice voice, the first slice voice is compared with the second ward-round voice, the slice voice which does not contain the voiceprint corresponding to the second ward-round voice is taken as second background noise, and the second background noise is removed from the slice voice which is closest in subsequent distance time and contains the voiceprint corresponding to the second ward-round voice to obtain a second slice voice;
Specifically, since background noise may be different at different times in the process of obtaining the third ward-round voice, in order to obtain background noise in real time and obtain accurate fourth ward-round voice, the third ward-round voice is sliced, and a plurality of first sliced voices are obtained, and since there is a pause between sentences when the attending doctor evaluates the attenuation risk of the patient, the first sliced voice that does not include voiceprint in the second ward-round voice, i.e., the voiceprint of the attending doctor, is the second background noise, and denoising processing is performed on the first sliced voice before the next second background noise after the second background noise, so as to obtain the second sliced voice, thereby laying a foundation for accurately obtaining ward-round information.
Step S24: and splicing the second sliced voice to obtain fourth ward-round voice, converting the fourth ward-round voice into second text information through the semantic conversion unit, and extracting the ward-round information from words analyzed by the second text information.
Specifically, the denoised second sliced speech is spliced to obtain pure fourth ward-round speech, the fourth ward-round speech is converted into the second text information through the semantic conversion unit, words related to the attenuation risk of the patient are obtained from the words of the second text information through semantic analysis, and the content of the words related to the attenuation risk of the patient is obtained through sentences in the second text information and grammar structures of the sentences, so that accurate ward-round information is obtained, and further more accurate attenuation risk levels are obtained.
Further, the step S3 includes the following steps:
Step S31: acquiring a test image template matched with the test image based on the distribution of tables, characters, numbers and figures in the test image and the area occupied by the test image;
Step S32: dividing the inspection image into a plurality of sub-inspection images based on the position information of each inspection object in the inspection image template, comparing each sub-inspection image with each reference image, and obtaining a plurality of target images;
Step S33: and converting a plurality of the target images into a test text by a conversion unit, and extracting the test data from the test text.
Specifically, since the format and content corresponding to different inspection images are different, for example: the method comprises the steps of obtaining a reference object of the inspection image from the storage unit, wherein the reference object can be a table format and distribution positions of characters, numbers and patterns in the inspection image, obtaining position information corresponding to each inspection item based on the inspection image template, dividing the inspection image into a plurality of target images based on the position information, converting the targets into inspection texts, extracting inspection data from the inspection texts, and obtaining accurate inspection data by the technical scheme and providing basis for obtaining the risk level of a patient with accurate identification precision of the monitoring image.
Further, the step S4 includes:
Comparing the test data with recent historical test data, judging a first risk change of the patient according to a comparison result, comparing the first risk change with a second risk change obtained by comparing last ward round information with current ward round information, and obtaining a comparison result, when the comparison result is consistent with the first risk change and the second risk change, increasing the identification precision of the monitoring image to be first precision when the first risk change and the second risk change are attenuation risks and rise, and decreasing the identification precision of the monitoring image to be second precision when the first risk change and the second risk change are attenuation risks and fall, and when the first risk change and the second risk change are attenuation risks and invariable, the identification precision of the monitoring image is invariable, wherein the first precision is larger than the second precision;
And carrying out set proportion amplification on the monitoring image and the reference monitoring image, wherein the higher the identification precision of the monitoring image is, the larger the corresponding set proportion is, the grid division is carried out on the monitoring image and the reference monitoring image, the grid comparison is carried out on the monitoring image and the reference monitoring image one by one, the first image closest to the monitoring image in the reference monitoring image is obtained, and the monitoring information corresponding to the first image is obtained.
Specifically, the first risk change is reduced by comparing the inspection data with the latest history inspection data stored in the storage unit, when the comparison result is that the inspection data is superior to the history inspection data, whereas the first risk change is increased when the comparison result is that the inspection data is inferior to or equal to the history inspection data, wherein the superiority or inferiority between different inspection data and the history inspection data is judged by comparing the inspection data with the inspection standard of each inspection item in the history inspection data, and the inspection data corresponding to each inspection item is compared according to the inspection standard corresponding to each inspection item, thereby obtaining the attenuation risk change, and the second risk change is also obtained from the ward information, wherein the ward information is evaluation information of the attenuation risk change of the patient, and since the ward and the body fluid, the blood or other examination items related to the viscera of the patient are required to be examined every day by the serious patient, the first risk change and the second risk change are both risk changes relative to the previous day, and the recognition accuracy of the monitoring image is adjusted by the first risk change and the second risk change, and when the first risk change and the second risk change are both attenuation risk increases, serious consequences may be caused due to the increase of the attenuation risk level of the patient, and since the monitoring image can reflect the real-time state of the patient, the recognition accuracy of the monitoring image needs to be increased to timely recognize the attenuation change and the change amount of the patient, on the contrary, when the first risk change and the second risk change are both attenuation risks, the recognition accuracy of the monitoring images is slightly reduced, the data processing workload in the monitoring image recognition process is reduced, so that monitoring images of other patients with larger attenuation risks can be processed in time, the monitoring images are amplified based on the recognition accuracy and compared with corresponding reference monitoring images, a first image closest to the monitoring images is obtained, the reference monitoring images are standard monitoring images of the patient with different attenuation degrees of organs of the patient, and monitoring information corresponding to each standard monitoring image is stored in a storage unit and used for reflecting the attenuation degree of the organs of the patient.
Further, the step S4 further includes: and when the first risk change is inconsistent with the second risk change, acquiring a first average value and a second average value of the monitoring information of the previous day and today, comparing the first average value with the second average value, acquiring a third risk change, when the second risk change is inconsistent with the third risk change, acquiring the ward-round information again through the stored third ward-round voice and the method of the step S2, when the first risk change is inconsistent with the third risk change, acquiring the inspection data again through the step S3, and setting the recognition precision of the monitoring image according to the third risk change.
Specifically, when the first risk change and the second risk change are different, that is, one of the first risk change and the second risk change indicates that the risk of the reduction of the organ of the patient is reduced, and the other indicates that the risk of the reduction of the organ of the patient is increased, possibly the first risk change or the second risk change is incorrect, therefore, the identification accuracy of the monitoring image is set to a default accuracy, monitoring information is acquired through the monitoring image based on the default accuracy, the third risk change is acquired by comparing the first average value and the second average value of the monitoring information of the previous day and the current day, wherein the inspection data or the ward information corresponding to the first risk change or the second risk change inconsistent with the third risk change is acquired in error, so that the inspection data or the ward information is re-acquired through the step S2 or the step S3, thereby ensuring the accuracy of the inspection data and the ward information, and also ensuring that when the patient is increased in the risk of the monitoring image is acquired, the accurate monitoring information is acquired, and the work load of processing the image is reduced when the patient is reduced in the risk of the patient is reduced.
Further, the step S5 includes: and scoring the monitoring information, the ward-round information and the inspection data through a SOFA scoring rule, determining the risk level based on the scoring, and carrying out early warning according to the risk level.
Specifically, all the SOFA scoring items can be obtained through the monitoring information, the ward-round information and the inspection data, and the real-time monitoring information can be further obtained through the monitoring image obtained in real time, so that not only the accurate SOFA scoring but also the SOFA scoring can be obtained in real time through the technical scheme, wherein the scoring is performed through the SOFA scoring table as the prior art, and the risk level is determined based on the scoring, for example: the first numerical value is divided into a first level, the second numerical value is divided into a second level, the third numerical value is divided into a third level, and the third numerical value is divided into a fourth level, wherein the first numerical value, the second numerical value and the third numerical value are increased once, early warning is carried out based on the risk levels, corresponding measures are taken in time, and deterioration of visceral organ attenuation of a patient is prevented.
Further, both the first risk change and the second risk change are attenuation risk changes for the patient on the same day relative to attenuation risk changes for the previous day.
Further, the ward-round information includes SOFA scoring item information in addition to the inspection data and the monitoring information.
The invention also provides a system for determining attenuation risk level and carrying out early warning through SOFA score, which is used for realizing the method, as shown in fig. 2, and comprises the following steps:
the acquisition unit is used for acquiring a monitoring image of a user in real time, acquiring a checking image and first ward-round voice in time, and preprocessing the monitoring image and the checking image;
the processing unit is configured to: denoising the first ward-round voice to obtain a second ward-round voice, obtaining first text information by the second ward-round voice through a semantic conversion unit, extracting patient information from the first text information, continuously receiving a third ward-round voice, and obtaining a fourth ward-round voice through denoising; the method comprises the steps of comparing the monitoring image with a reference monitoring image according to ward-round information and inspection data to obtain a first image closest to the monitoring image, and obtaining monitoring information of the first image; extracting inspection data from the inspection image;
the semantic conversion unit is used for converting the fourth ward-round voice into second text information, acquiring ward-round information from the second text information and storing the third ward-round voice;
And the early warning unit is used for scoring according to the monitoring information, the ward-round information and the inspection data to determine the risk level and carrying out early warning based on the risk level.
In summary, the present invention obtains the first ward round voice in a ward round preset time period, obtains the second ward round voice through the first ward round voice denoising process, recognizes the patient identification information through converting the second ward round voice into the first text information, further continues to obtain the third ward round voice, slices the third ward round voice, recognizes the second background noise, denoises the first slice voice, obtains the second slice voice, splices the second slice voice to obtain the fourth ward round voice, obtains the ward round information through the second text information, obtains the inspection image through the inspection image, obtains the first risk change through comparing the inspection data of the previous day of the patient with the inspection data of the current day, obtains the second risk change through comparing the ward round information of the current day with the first ward round information, obtains the second risk change based on the first risk change with the first ward round information, obtains the second risk change, obtains the image attenuation corresponding to the first risk change, and obtains the image attenuation corresponding to the image attenuation of the patient according to the first image, and can further reduce the image attenuation corresponding to the image attenuation of the first image attenuation, and can further obtain the image attenuation corresponding to the image attenuation.
The technical features of the above embodiments may be arbitrarily combined, and for brevity, all of the possible combinations of the technical features of the above embodiments are not described, however, they should be considered as the scope of the description of the present specification as long as there is no contradiction between the combinations of the technical features.
The foregoing examples have been presented to illustrate only a few embodiments of the invention and are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (8)
1. A method for determining risk level and pre-warning by attenuation of SOFA score, the method comprising the steps of:
Step S1: acquiring a monitoring image of a user through an acquisition unit in real time, acquiring a checking image and first ward-round voice in time, and preprocessing the monitoring image and the checking image;
Step S2: denoising the first ward-round voice to obtain second ward-round voice, obtaining first text information from the second ward-round voice through a semantic conversion unit, extracting patient information from the first text information, continuously receiving third ward-round voice, obtaining fourth ward-round voice through denoising, converting the fourth ward-round voice into second text information through the semantic conversion unit, obtaining ward-round information from the second text information, and storing the third ward-round voice;
step S3: extracting test data from the test image by an extraction unit;
Step S4: according to the ward-round information and the inspection data, comparing the first image closest to the monitoring image based on the monitoring image and a reference monitoring image, and acquiring monitoring information of the first image;
step S5: scoring according to the monitoring information, the ward-round information and the inspection data to determine a risk level and performing early warning based on the risk level;
Wherein, the step S2 includes: step S21: periodically acquiring the first ward-round voice in a preset time period through the extraction unit, taking the first ward-round voice acquired last time as first background noise, and preprocessing the first ward-round voice based on the first background noise to acquire the second ward-round voice with the first background noise removed;
Step S22: converting the second ward-round voice into first text information through a semantic conversion module, analyzing the text information into words, storing the second ward-round voice when the words of the first text information comprise identification information of a patient, and reducing the acquisition period of the first ward-round voice;
Step S23: the first ward-round voice is continuously and periodically obtained to serve as a third ward-round voice, the third ward-round voice is sliced to obtain a first slice voice, the first slice voice is compared with the second ward-round voice, the slice voice which does not contain the voiceprint corresponding to the second ward-round voice is taken as second background noise, and the second background noise is removed from the slice voice which is closest in subsequent distance time and contains the voiceprint corresponding to the second ward-round voice to obtain a second slice voice;
step S24: splicing the second sliced voice, acquiring second text information through the semantic conversion unit, and extracting ward round information from words analyzed by the second text information;
The step S4 includes:
Comparing the test data with recent historical test data, judging a first risk change of a patient according to a comparison result, comparing the first risk change with a second risk change obtained by comparing last ward round information with current ward round information, and obtaining a comparison result, when the comparison result is consistent with the first risk change and the second risk change, increasing the identification precision of the monitoring image to be first precision when the first risk change and the second risk change are both attenuation risk increase, reducing the identification precision of the monitoring image to be second precision when the first risk change and the second risk change are both attenuation risk decrease, and keeping the identification precision of the monitoring image unchanged when the first risk change and the second risk change are both attenuation risk unchanged, wherein the first precision is larger than the second precision;
And carrying out set proportion amplification on the monitoring image and the reference monitoring image, wherein the higher the identification precision of the monitoring image is, the larger the corresponding set proportion is, carrying out grid division on the monitoring image and the reference monitoring image, carrying out grid-by-grid comparison on the monitoring image and the reference monitoring image one by one, obtaining the first image closest to the monitoring image in the reference monitoring image, and obtaining monitoring information corresponding to the first image.
2. The method of claim 1, wherein the monitoring image is a display image of a monitoring result obtained by monitoring by a monitoring device, the inspection image is an inspection result image, and the first ward-round voice is an evaluation dialogue for patient's condition development during ward-round by a doctor.
3. The method according to claim 1, wherein in the step S1, the monitoring image and the inspection image are preprocessed: performing angle adjustment and binarization on the monitoring image; the inspection image is enhanced.
4. The method according to claim 1, wherein the step S3 comprises the steps of:
Step S31: acquiring a test image template matched with the test image based on the distribution of tables, characters, numbers and figures in the test image and the area occupied by the test image;
step S32: dividing the inspection image into a plurality of sub-inspection images based on the inspection image template, comparing each sub-inspection image with each reference image, and obtaining a plurality of target images;
Step S33: and converting a plurality of the target images into a test text by a conversion unit, and extracting the test data from the test text.
5. The method according to claim 1, wherein the step S4 further comprises: and when the first risk change is inconsistent with the second risk change, acquiring a first average value and a second average value of the monitoring information of the previous day and today, comparing the first average value with the second average value, acquiring a third risk change, when the second risk change is inconsistent with the third risk change, acquiring the ward-round information again through the stored third ward-round voice and the method of the step S2, when the first risk change is inconsistent with the third risk change, acquiring the inspection data again through the step S3, and setting the recognition precision of the monitoring image according to the third risk change.
6. The method according to claim 1, wherein the step S5 comprises: and scoring the monitoring information, the ward-round information and the inspection data through a SOFA scoring rule, determining the risk level based on the scoring, and carrying out early warning according to the risk level.
7. The method of claim 1, wherein the first risk change and the second risk change are both decay risk changes for a day of the patient relative to a day before, the ward information including SOFA scoring information in addition to the test data and the monitoring information.
8. A system for determining a risk level of attenuation and early warning by SOFA scoring, characterized in that the system is adapted to implement the method of any of claims 1-7, the system comprising:
the acquisition unit is used for acquiring a monitoring image of a user in real time, acquiring a checking image and first ward-round voice in time, and preprocessing the monitoring image and the checking image;
the processing unit is configured to: denoising the first ward-round voice to obtain a second ward-round voice, obtaining first text information by the second ward-round voice through a semantic conversion unit, extracting patient information from the first text information, continuously receiving a third ward-round voice, and obtaining a fourth ward-round voice through denoising; the method comprises the steps of comparing the monitoring image with a reference monitoring image according to ward-round information and inspection data to obtain a first image closest to the monitoring image, and obtaining monitoring information of the first image; extracting inspection data from the inspection image;
the semantic conversion unit is used for converting the fourth ward-round voice into second text information, acquiring ward-round information from the second text information and storing the third ward-round voice;
And the early warning unit is used for scoring according to the monitoring information, the ward-round information and the inspection data to determine the risk level and carrying out early warning based on the risk level.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410591821.0A CN118173272B (en) | 2024-05-14 | 2024-05-14 | Method for determining risk level and carrying out early warning through attenuation of SOFA score |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410591821.0A CN118173272B (en) | 2024-05-14 | 2024-05-14 | Method for determining risk level and carrying out early warning through attenuation of SOFA score |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118173272A CN118173272A (en) | 2024-06-11 |
| CN118173272B true CN118173272B (en) | 2024-08-02 |
Family
ID=91360856
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410591821.0A Active CN118173272B (en) | 2024-05-14 | 2024-05-14 | Method for determining risk level and carrying out early warning through attenuation of SOFA score |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118173272B (en) |
Family Cites Families (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012051394A1 (en) * | 2010-10-14 | 2012-04-19 | The Arizona Board Of Regents On Behalf Of The University Of Arizona | Methods and apparatus for imaging detecting, and monitoring surficial and subdermal inflammation |
| US20150213217A1 (en) * | 2012-09-13 | 2015-07-30 | Parkland Center For Clinical Innovation | Holistic hospital patient care and management system and method for telemedicine |
| CN110111778B (en) * | 2019-04-30 | 2021-11-12 | 北京大米科技有限公司 | Voice processing method and device, storage medium and electronic equipment |
| CN112151172A (en) * | 2019-06-28 | 2020-12-29 | 希尔-罗姆服务公司 | Patient risk assessment based on data from multiple sources in a medical facility |
| CN110459328B (en) * | 2019-07-05 | 2020-07-28 | 浙江大学 | Clinical monitoring equipment |
| CN110634557B (en) * | 2019-08-23 | 2022-08-23 | 广东省智能制造研究所 | Medical care resource auxiliary allocation method and system based on deep neural network |
| CN112786199B (en) * | 2019-11-05 | 2024-06-25 | 希尔-罗姆服务公司 | Interface for displaying patient data |
| AU2021264555A1 (en) * | 2020-04-29 | 2022-11-17 | Inflammatix, Inc. | Determining mortality risk of subjects with viral infections |
| CN116052824A (en) * | 2021-10-28 | 2023-05-02 | 深圳市理邦精密仪器股份有限公司 | Ward round method, ward round equipment and system |
| CN114067340B (en) * | 2022-01-17 | 2022-05-20 | 山东北软华兴软件有限公司 | Intelligent judgment method and system for information importance |
| CN114882983B (en) * | 2022-05-13 | 2023-02-10 | 浙江远图技术股份有限公司 | Smart ward interaction method, system and storage medium |
| CN116543750A (en) * | 2023-04-03 | 2023-08-04 | 淮北矿业股份有限公司 | Precision lifting system for voice recognition |
| CN117393091A (en) * | 2023-09-08 | 2024-01-12 | 中国人民解放军总医院 | Ward round method and device, mobile robot and storage medium |
| CN117198446A (en) * | 2023-09-08 | 2023-12-08 | 厦门狄耐克物联智慧科技有限公司 | Ward-round report automatic generation method based on medical care intercom equipment |
| CN117877731A (en) * | 2024-01-16 | 2024-04-12 | 沈阳方晟医药科技有限公司 | Real-time risk early warning and monitoring system and method for cardiogenic shock |
| CN117612725B (en) * | 2024-01-23 | 2024-03-29 | 南通大学附属医院 | Respirator alarm management method and system for intensive care unit |
| CN117995373B (en) * | 2024-04-07 | 2024-06-21 | 北京惠每云科技有限公司 | Hospital ward record generation method and device, electronic equipment and storage medium |
-
2024
- 2024-05-14 CN CN202410591821.0A patent/CN118173272B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN118173272A (en) | 2024-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109859168A (en) | A kind of X-ray rabat picture quality determines method and device | |
| KR102676569B1 (en) | Medical image analysis apparatus and method, medical image visualization apparatus and method | |
| CN111274400B (en) | Construction method, device, equipment and storage medium of medical term system | |
| CN111275755B (en) | Mitral valve orifice area detection method, system and equipment based on artificial intelligence | |
| CN117542536B (en) | Intelligent nursing method and system based on physical examination data | |
| CN109102888A (en) | A kind of human health methods of marking | |
| CN117809798B (en) | Verification report interpretation method, system, equipment and medium based on large model | |
| US20240020284A1 (en) | Medical Clinical Data Quality Analysis System Based on Big Data | |
| CN116864104A (en) | Chronic thromboembolic pulmonary artery high-pressure risk classification system based on artificial intelligence | |
| CN112132800A (en) | Pulmonary fibrosis detection and severity evaluation method and system based on deep learning | |
| CN115762769A (en) | Intelligent risk early warning system after ERCP (effective Range planning) operation | |
| CN114343577A (en) | Cognitive function evaluation method, terminal device, and computer-readable storage medium | |
| EP3939003A1 (en) | Systems and methods for assessing a likelihood of cteph and identifying characteristics indicative thereof | |
| CN118173272B (en) | Method for determining risk level and carrying out early warning through attenuation of SOFA score | |
| CN113974578A (en) | Device for estimating physiological heart measurement according to qi and blood analysis | |
| CN112259232B (en) | VTE risk automatic evaluation system based on deep learning | |
| US20030191732A1 (en) | Online learning method in a decision system | |
| CN109002838B (en) | IVUS vulnerable plaque tissue classification method and system based on vulnerable index | |
| CN116597993A (en) | Hypertension risk dynamic assessment system for pregnant women in gestation period | |
| EP3443492A1 (en) | Annotating data points associated with clinical decision support application | |
| US20230293103A1 (en) | Analysis device | |
| CN110910980A (en) | Sepsis early warning device, equipment and storage medium | |
| CN117558391B (en) | Postoperative condition deep learning method and system based on aortic dissection medical record data | |
| CN111297348A (en) | Method for generating electrocardiogram analysis result, electrocardiograph and storage medium | |
| CN118098483B (en) | Medical record writing monitoring method and device based on transducer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |