Medical decision weight distribution method based on AI

Abstract

The invention relates to the field of intelligent medical treatment, in particular to an AI-based medical decision weight distribution method, which comprises the steps of determining a data extraction mode according to data richness; during multi-data source extension selection, determining a data matching degree according to a historical morbidity and a rhythm disorder coefficient, and taking data with the data matching degree greater than a preset data matching degree as to-be-supplemented data; determining a weight coefficient of a related parameter of the data matching degree according to the critical threshold value; according to the effective data relevancy, judging that the supplement selection mode is to select the medical record data according to the feature keywords or select the medical record data according to the dynamic difference coefficient; determining a weight adjustment amplitude according to a reference weight distribution mode or a diagnosis and treatment validity difference value according to a weight adjustment mode corresponding to the effective data relevancy determined according to the diagnosis and treatment stability coefficient; when the weight adjustment amplitude value reaches a preset adjustment amplitude value, weight adjustment is ended; and the accuracy and efficiency of medical record file matching are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of smart medical treatment, in particular to a medical decision weight distribution method based on AI. BACKGROUND

[0002] The existing medical decision weight distribution technology based on deep learning fails to determine a targeted processing method according to the feature state of the medical record in the actual matching process, thereby failing to select a matching method in a targeted manner in the subsequent medical record matching process, resulting in low accuracy and processing efficiency of medical record matching.

[0003] Chinese Patent Publication No. CN119230108A discloses a medical decision assistance method and system based on medical rules, which includes receiving medical data of a patient from various medical devices and information systems, standardizing the received data, mapping the standardized data to corresponding medical concepts after data standardization, and obtaining and recording the associated information of the data during data mapping. The present application effectively solves the misassociation problem of different source medical data in association through standardization processing and machine learning model analysis, improves the accuracy of diagnosis and treatment, and prevents misdiagnosis and over-treatment by evaluating data association quality and screening out misassociated data. The clustering algorithm is used to optimize abnormal data identification and exclusion, ensure reliable data-based diagnosis, reduce unnecessary medical intervention caused by data errors, and improve overall nursing quality and medical service level. It can be seen that the above technical solution has the following problems: it fails to determine a targeted processing method according to the feature state of the medical record in the actual matching process, thereby failing to select a matching method in a targeted manner in the subsequent medical record matching process, resulting in low accuracy and processing efficiency of medical record matching. SUMMARY

[0004] Therefore, the present application provides a medical decision weight distribution method based on AI to overcome the problem that the prior art fails to determine a targeted processing method according to the feature state of the medical record in the actual matching process, thereby failing to select a matching method in a targeted manner in the subsequent medical record matching process, resulting in low accuracy and processing efficiency of medical record matching.

[0005] To achieve the above-mentioned purpose, the present application provides a medical decision weight distribution method based on AI, which comprises: determining the data extraction method as multi-data source extension selection or data supplement according to the data richness; when multi-data source extension selection is selected, the data matching degree is determined according to the historical incidence rate and the rhythm disorder coefficient, and the data with a data matching degree greater than a preset data matching degree is selected as the data to be supplemented; The weight coefficient of the related parameter of the data matching degree is determined according to the critical threshold value; and the weight of any related parameter greater than the critical threshold value is increased and adjusted, and the increase adjustment range is determined according to the effective difference value; the related parameter is a historical incidence rate difference value or a rhythm disorder coefficient difference value; the difference value of the historical incidence rate of the target medical record and the comparison medical record is recorded as the historical incidence rate difference value; and the difference value of the rhythm disorder coefficient of the target medical record and the comparison medical record is recorded as the rhythm disorder coefficient difference value.

[0006] According to the comparison result of the effective data correlation degree and the preset effective data correlation degree, it is determined that the supplement selection mode is to select medical record data according to the feature keyword or to select medical record data according to the dynamic difference coefficient; According to the comparison result of the diagnosis and treatment stability coefficient and the preset diagnosis and treatment stability coefficient, the weight adjustment mode corresponding to the effective data correlation degree is determined to be according to the reference weight distribution mode or to determine the weight adjustment range according to the diagnosis and treatment effective degree difference value; The data source includes medical records, laboratory data and image reports; When the weight adjustment range reaches the preset adjustment range, the weight adjustment is ended.

[0007] Further, the data extraction mode is determined according to the data richness, and the extraction mode includes: If the data richness is greater than the preset data richness, the data extraction mode is multi-data source extension selection; If the data richness is less than or equal to the preset data richness, the data extraction mode is data supplement according to the effective correlation degree.

[0008] Further, in response to the first preset supplement condition, the data matching degree is determined according to the historical incidence rate and the rhythm disorder coefficient, and the data with a data matching degree greater than a preset data matching degree is taken as the data to be supplemented; The first preset supplement condition is to determine that the data extraction mode is multi-data source extension selection.

[0009] Further, the weight coefficient of the related parameter of the data matching degree is determined according to the critical threshold value; and the weight of any related parameter greater than the critical threshold value is increased and adjusted, and the increase adjustment range is determined according to the effective difference value; The related parameter is a historical incidence rate difference value or a rhythm disorder coefficient difference value; the difference value of the historical incidence rate of the target medical record and the comparison medical record is recorded as the historical incidence rate difference value; and the difference value of the rhythm disorder coefficient of the target medical record and the comparison medical record is recorded as the rhythm disorder coefficient difference value.

[0010] Further, in response to the second preset supplement condition, the supplement selection mode is determined according to the comparison result of the effective data correlation degree and the preset effective data correlation degree; If the effective data correlation degree is less than the preset effective data correlation degree, it is determined that the supplement selection mode is to select medical record data according to the feature keyword. If the effective data correlation degree is greater than or equal to a preset effective data correlation degree, it is determined that the supplementary selection mode is selected according to the dynamic difference coefficient; The second preset supplementary condition is to determine the data extraction mode as data supplement according to the effective correlation degree.

[0011] Further, the effective data correlation degree is determined according to the keyword repetition degree and the keyword proportion; The effective data correlation degree is positively correlated with the keyword repetition degree and the keyword proportion, respectively.

[0012] Further, the effective data correlation degree is determined according to the weight adjustment mode corresponding to the effective data correlation degree; If the treatment effective coefficient is greater than a preset treatment effective coefficient, the reference weight distribution mode is used; If the treatment effective coefficient is less than or equal to the preset treatment effective coefficient, the weight adjustment amplitude is determined according to the treatment effective degree difference; The adjustment amplitude is positively correlated with the treatment effective degree difference.

[0013] Further, the treatment effective coefficient is determined according to the cure rate and the recurrence rate; The treatment effective coefficient is positively correlated with the cure rate; The treatment effective coefficient is negatively correlated with the recurrence rate.

[0014] Further, the data source includes medical records, laboratory data and image reports.

[0015] Further, when the weight adjustment amplitude reaches a preset adjustment amplitude, the weight adjustment is ended.

[0016] Compared with the prior art, the beneficial effects of the present application are that in the technical scheme of the present application, two kinds of data extraction modes are determined according to the comparison result of the data richness degree and the preset data richness degree, which avoids the problem of over-extraction or missing of key data caused by a single data extraction mode in the prior art, and further improves the accuracy and efficiency of data collection.

[0017] Further, in the technical scheme of the present application, the data matching degree is determined by the historical incidence rate and the rhythm disorder coefficient, which avoids the problem that a single influencing factor cannot accurately determine the data matching degree, improves the accuracy of the data matching degree, and is further beneficial to the subsequent adjustment of the weight, and further improves the accuracy and processing efficiency of medical record matching.

[0018] Further, the application determines the supplement selection mode according to the comparison result of the effective data correlation degree and the preset effective data correlation degree, avoids the single supplement selection mode from being difficult to meet the actual application requirement, and further improves the rationality of the supplement data, and determines the weight adjustment mode corresponding to the effective data correlation degree according to the diagnosis and treatment stability coefficient, and further improves the rationality of the weight adjustment, and further improves the accuracy and processing efficiency of the medical record matching. BRIEF DESCRIPTION OF DRAWINGS

[0019] Fig. 1 A schematic diagram of the AI-based medical decision weight distribution method of the application; Fig. 2 A flowchart of determining the data extraction mode according to the data richness of the application; Fig. 3 A flowchart of determining the supplement selection mode according to the effective data correlation degree of the application. DETAILED DESCRIPTION

[0020] In order to make the objects and advantages of the application clearer, the application will be further described below with reference to the embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and do not limit the application.

[0021] The preferred embodiments of the application will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the application, and are not intended to limit the protection scope of the application.

[0022] It should be noted that in the description of the application, the terms "upper", "lower", "left", "right", "inner", "outer" and the like indicate the direction or positional relationship of the terms based on the direction or positional relationship shown in the drawings, which is only for the convenience of description, and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the application.

[0023] In addition, it should be noted that in the description of the application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication inside two elements. Those skilled in the art can understand the specific meaning of the above terms in the application according to the specific circumstances.

[0024] Please refer to Figs. 1 to 3 The application provides an AI-based medical decision weight distribution method, which comprises: According to the data richness, the data extraction mode is determined as multi-data source extension selection or data supplement according to effective correlation; In the multi-data source extension selection, the data matching degree is determined according to the historical incidence rate and the rhythm disorder coefficient, and the data with a data matching degree greater than a preset data matching degree is taken as the data to be supplemented; According to the critical threshold, the weight coefficient of the related parameter of the data matching degree is determined; the weight of any related parameter greater than the critical threshold is adjusted by increasing, and the increasing adjustment amplitude is determined according to the effective difference value; Among them, the related parameters are historical incidence rate difference or rhythm disorder coefficient difference, the difference between the historical incidence rate of the target medical record and the comparison medical record is recorded as the historical incidence rate difference, and the difference between the rhythm disorder coefficient of the target medical record and the comparison medical record is recorded as the rhythm disorder coefficient difference.

[0025] According to the comparison result of the effective data correlation degree and the preset effective data correlation degree, it is determined that the supplement selection mode is to select medical record data according to feature keywords or to select medical record data according to dynamic difference coefficient; According to the comparison result of the diagnosis and treatment stability coefficient and the preset diagnosis and treatment stability coefficient, the weight adjustment mode corresponding to the effective data correlation degree is determined as the reference weight distribution mode or the weight adjustment amplitude is determined according to the diagnosis and treatment effective degree difference; The data source includes medical records, laboratory data and image reports; When the weight adjustment amplitude reaches the preset adjustment amplitude, the weight adjustment is ended.

[0026] The application scenario of the present application is to screen and match medical records for medical decision-making by AI, so as to improve the efficiency of auxiliary decision-making.

[0027] Specifically, the data extraction mode is determined according to the data richness, and the extraction mode includes: If the data richness is greater than the preset data richness, the data extraction mode is multi-data source extension selection; If the data richness is less than or equal to the preset data richness, the data extraction mode is data supplement according to effective correlation.

[0028] The confirmation method of the data richness is that, for a target disease, the number of patients with the target disease is detected, and the data richness = the number of patients with the target disease / the total population; The value of the preset data richness can be adaptively set by the user according to the actual application scenario. It can be understood that the greater the value of the preset data richness, the greater the demand of the user for multi-data source extension selection. A value of the preset data richness is provided, the historical records of the user adopting multi-data source extension selection are detected, and the average value of the data richness corresponding to the historical records meeting the demand of the user is taken as the preset data richness.

[0029] Specifically, in response to the first preset supplement condition, the data matching degree is determined according to the historical incidence rate and the rhythm disorder coefficient, and the data with a data matching degree greater than a preset data matching degree is taken as the to-be-supplemented data. The first preset supplement condition is that the data extraction mode is determined to be multi-data source extension selection.

[0030] The data matching degree = a1 x historical incidence rate + a2 x rhythm disorder coefficient; wherein a1 is a first weight coefficient, a2 is a second weight coefficient, the values of a1 and a2 can be directly set by the user according to field experience, or the contribution of the historical incidence rate and the rhythm disorder coefficient to the data matching degree is determined by using a statistical method such as regression analysis or principal component analysis (PCA), so as to determine the values of the corresponding weight coefficients, the greater the contribution, the greater the value of the weight coefficient, and the weight can also be adjusted by historical records (such as machine learning).

[0031] The number of disease occurrences of a plurality of patients with the target disease is detected, wherein for a single patient with the target disease, the number of disease occurrences of the patient from the time of initial diagnosis of the target disease to the time of the most recent visit is detected, and the number of disease occurrences / interval duration is recorded as the incidence rate of the patient, wherein the interval duration is the time of the most recent visit-the time of initial diagnosis of the target disease, and the average value of the incidence rates of the plurality of patients is recorded as the historical incidence rate, wherein the value of the number of patients can be adaptively set by the user according to actual requirements, and it can be understood that the higher the accuracy requirement of the user for the historical incidence rate, the greater the value of the number of patients, and the present application provides a value of the number of patients, which is 500 in the present application.

[0032] The rhythm disorder coefficient = |peak coefficient corresponding to the initial visit time of the target disease-preset peak coefficient|, and the preset peak coefficient is confirmed in the following manner: for a target reference value corresponding to the initial visit time of the target disease, the target reference value is recorded as a1, and the average value of the peak coefficients corresponding to the target reference value data with a1 in the historical records is recorded as the preset peak coefficient; the peak coefficients corresponding to different diseases are different, and the user can adaptively set them according to actual application requirements.

[0033] The value of the preset data matching degree can be adaptively set by the user according to actual application requirements, and it can be understood that the higher the matching degree requirement of the user for the to-be-supplemented data, the greater the value of the preset data matching degree, and the present application provides a value of the preset data matching degree, which is determined by detecting the average value of the data matching degrees in the historical records that meet the user's requirements.

[0034] Specifically, the weight coefficient of the related parameter of the data matching degree is determined according to the critical threshold value; and the weight of any related parameter greater than the critical threshold value is adjusted to increase, and the increase adjustment range is determined according to the effective difference value. The related parameter is a historical incidence difference value or a rhythm disorder coefficient difference value, the difference value between the historical incidence of the target medical record and the comparison medical record is referred to as the historical incidence difference value, and the difference value between the rhythm disorder coefficient of the target medical record and the comparison medical record is referred to as the rhythm disorder coefficient difference value.

[0035] The value of the critical threshold value can be adaptively set by the user according to actual application requirements. It can be understood that the higher the precision requirement of the user for the data matching degree, the smaller the value of the critical threshold value. The application provides a value setting method of the critical threshold value, extracts the corresponding threshold value in the historical record meeting the user's requirements, screens out the abnormal values, and takes the average value of the threshold values after removing the abnormal values as the critical threshold value.

[0036] The effective difference value is the larger one of the historical incidence difference value and the rhythm disorder coefficient difference value. The weight of the effective difference value greater than the critical threshold value is adjusted to increase on the basis of the basic weight. In the application, the basic weights are α1=0.5 and α2=0.5, the increase amplitude=adjustment coefficient×(effective difference value-critical threshold value); the value of the adjustment coefficient can be adaptively set by the user according to actual application requirements. It can be understood that the greater the difference between the effective difference value and the critical threshold value, the greater the influence on the increase amplitude, and the greater the value of the adjustment coefficient. The application provides a value setting method of the adjustment coefficient, and the adjustment coefficient in the application is 0.5.

[0037] Specifically, in response to the second preset supplement condition, the supplement selection method is determined according to the comparison result of the effective data correlation degree and the preset effective data correlation degree. If the effective data correlation degree is less than the preset effective data correlation degree, it is determined that the supplement selection method is to select medical record data according to the feature keyword. If the effective data correlation degree is greater than or equal to the preset effective data correlation degree, it is determined that the supplement selection method is to select medical record data according to the dynamic difference coefficient. The second preset supplement condition is that the data extraction method is determined to be data supplement according to the effective correlation degree.

[0038] The key words in the medical record are identified through machine learning. The data meeting the format of "key word:value" is recorded as a combination, and the combination is recorded as a feature keyword. It can be understood that the body temperature in the medical record information is a key word, and the body temperature: 36.5℃ is a feature keyword.

[0039] The data in which the feature keywords exist in the medical record scheme is recorded as one type of selection data, and the one type of selection data is selected in descending order of the number of feature keywords, until the number of selected one type of selection data reaches a preset number; the number of feature keywords is the number of feature keywords contained in a single medical record data; the medical record data is selected according to the dynamic difference coefficient, including: the medical record data with a dynamic difference coefficient greater than a preset dynamic difference coefficient from the target medical record is recorded as two types of selection data, and the two types of selection data are selected in descending order of the dynamic difference coefficient, until the number of selected two types of selection data reaches a preset number; the value of the preset number can be adaptively set by the user according to the actual application scene, and the application provides a value setting method of the preset number, extracts the corresponding number in the historical record meeting the user's demand, screens out the outliers, and records the average value of the number after removing the outliers as the preset number.

[0040] The dynamic difference coefficient of the single matching medical record data and the target medical record data is the average value of the fluctuation difference values of each feature keyword corresponding to the matching medical record data and the target medical record data, and the fluctuation difference value of each feature keyword corresponding to the matching medical record data and the target medical record data is | the diagnosis and treatment fluctuation value of the feature keyword corresponding to the matching data - the diagnosis and treatment fluctuation value of the feature keyword corresponding to the target medical record |.

[0041] The diagnosis and treatment fluctuation value = standard deviation of parameter value at the visit time of the single matching data / drop frequency proportion, the drop frequency proportion = number of drop visit times / (number of visit times in the single matching data - 1), Wherein, different diseases correspond to different parameter values, and the parameter values include but are not limited to heart rate, blood sugar and tumor size.

[0042] The value of the preset effective data correlation degree can be adaptively set by the user according to the actual application requirement, and it can be understood that the user determines two different supplementary selection methods according to the comparison result of the effective data correlation degree and the preset effective data correlation degree, and the application provides a value setting method of the preset effective data correlation degree, extracts the corresponding effective data correlation degree in the historical record meeting the user's demand, screens out the outliers, and records the average value of the effective data correlation degree after removing the outliers as the preset effective data correlation degree.

[0043] Specifically, the effective data correlation degree is determined according to the keyword repetition degree and the keyword proportion. The effective data correlation degree is positively correlated with the keyword repetition degree and the keyword proportion, respectively.

[0044] The effective data correlation degree is equal to β1*keyword repetition degree+β2*keyword proportion, wherein β1 is a first weight coefficient, β2 is a second weight coefficient, the keyword repetition degree is determined according to the number of same keywords between the target medical record data and any matched medical record data, and the keyword repetition degree is equal to the number of same keywords / the number of keywords of the target medical record data; and the keyword proportion is equal to the number of keywords of the matched medical record data / the number of characters of the matched medical record data.

[0045] Specifically, the weight adjustment mode corresponding to the effective data correlation degree is determined according to the diagnosis and treatment effective coefficient; If the diagnosis and treatment effective coefficient is greater than the preset diagnosis and treatment effective coefficient, the weight adjustment mode is determined according to the reference weight distribution mode; If the diagnosis and treatment effective coefficient is less than or equal to the preset diagnosis and treatment effective coefficient, the weight adjustment range is determined according to the diagnosis and treatment effective degree difference; The adjustment range and the diagnosis and treatment effective degree difference are in a positive correlation relationship.

[0046] The diagnosis and treatment effective degree difference is equal to the preset diagnosis and treatment effective coefficient minus the diagnosis and treatment effective coefficient. The value of the preset diagnosis and treatment effective coefficient can be adaptively set by the user according to actual needs. It can be understood that the diagnosis and treatment effective coefficient reflects the diagnosis and treatment effect. When the diagnosis and treatment effective coefficient is greater than the preset diagnosis and treatment effective coefficient, the diagnosis and treatment effect is better, and therefore the weight distribution mode is not adjusted. In the present application, the value of the reference weight is provided, β1 is 0.5, and β2 is 0.5.

[0047] When the diagnosis and treatment effective coefficient is less than or equal to the preset diagnosis and treatment effective coefficient, the first weight coefficient is adjusted to increase, and the first weight coefficient is equal to 0.5 plus the increase range. The increase range is equal to the proportion coefficient*the diagnosis and treatment effective degree difference. The value of the proportion coefficient can be adaptively set by the user according to actual application needs. It can be understood that the greater the influence of the diagnosis and treatment effective degree difference on the increase range, the greater the value of the proportion coefficient. The ratio of the corresponding increase range to the diagnosis and treatment effective degree difference in the historical record meeting the user's needs is extracted, the abnormal values are screened out, and the average value of the ratio after removing the abnormal values is taken as the proportion coefficient.

[0048] It can be understood that the first weight coefficient+the second weight coefficient=1, and the increase range of the first weight coefficient is the same as the decrease range of the second weight coefficient.

[0049] Specifically, the diagnosis and treatment effective coefficient is determined according to the cure rate and the recurrence rate; The diagnosis and treatment effective coefficient and the cure rate are in a positive correlation relationship. The diagnosis and treatment effective coefficient and the recurrence rate are in a negative correlation relationship.

[0050] The treatment effective coefficient is equal to the cure rate plus 1 / the recurrence rate. When the key parameter data index and the compliance duration are greater than the preset duration in the medical record data of the target patient, the target patient is recorded as a cured patient. The cure rate is equal to the number of cured patients / the total number of patients. The probability of the cured patient suffering from the disease again within a preset recovery duration is recorded as the recurrence rate. The value of the preset recovery duration can be adaptively set by the user according to actual application requirements. It can be understood that different recovery durations are determined for different disease types. The recovery duration of a chronic disease is set to 1 year.

[0051] Specifically, the data sources include medical records, laboratory data, and image reports.

[0052] Specifically, when the weight adjustment amplitude reaches the preset adjustment amplitude, the weight adjustment is ended.

[0053] The value of the preset adjustment amplitude can be adaptively set by the user according to actual application requirements. The application provides a value setting method of the preset adjustment amplitude. The corresponding adjustment amplitudes in the historical records meeting the user's requirements are extracted, the abnormal values are screened out, and the average value of the adjustment amplitudes after removing the abnormal values is recorded as the preset adjustment amplitude. The application provides a value setting method of the preset adjustment amplitude. In the application, the preset adjustment amplitude is 60% of the basic weight.

[0054] So far, the technical solutions of the application have been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the application. The technical solutions after the changes or replacements will fall within the protection scope of the application.

Claims

1. An AI-based method for weighting medical decisions, characterized in that, include: The data extraction method is determined based on the data abundance: either by extending the selection from multiple data sources or by supplementing the data based on the effective relevance. When selecting multiple data sources, the data matching degree is determined based on historical incidence rate and rhythm disorder coefficient, and data with a matching degree greater than the preset matching degree are used as supplementary data. The weight coefficients of relevant parameters for data matching degree are determined based on the critical threshold; the weight of any relevant parameter that is greater than the critical threshold is increased and adjusted, and the adjustment range is determined based on the effective difference; the relevant parameters are the difference in historical incidence rate or the difference in rhythm disorder coefficient. The difference in historical incidence rate between the target medical record and the comparison medical record is recorded as the difference in historical incidence rate, and the difference in rhythm disorder coefficient between the target medical record and the comparison medical record is recorded as the difference in rhythm disorder coefficient. The supplementary selection method is determined based on the comparison results between the relevance of valid data and the preset relevance of valid data. The selection method is either based on the selection of medical record data according to the feature keywords or based on the dynamic difference coefficient. The weight adjustment method for the correlation of effective data is determined by comparing the diagnostic stability coefficient with the preset diagnostic stability coefficient. This method is either based on the baseline weight allocation method or by determining the weight adjustment magnitude based on the difference in diagnostic effectiveness. Data sources include medical records, laboratory data, and imaging reports; When the weight adjustment amplitude reaches the preset adjustment amplitude, the weight adjustment ends.

2. The AI-based medical decision-making weight allocation method according to claim 1, characterized in that, The data extraction method is determined based on data abundance. Extraction methods include: If the data abundance is greater than the preset data abundance, the data extraction method will be extended selection from multiple data sources; If the data abundance is less than or equal to the preset data abundance, the data extraction method is to supplement the data based on the effective relevance.

3. The AI-based medical decision-making weight allocation method according to claim 2, characterized in that, Under the first preset supplementation condition, the data matching degree is determined based on the historical incidence rate and the rhythm disorder coefficient, and the data with a matching degree greater than the preset data matching degree is regarded as data to be supplemented. The first preset supplementary condition is to determine that the data extraction method is a multi-source data source extended selection.

4. The AI-based medical decision-making weight allocation method according to claim 3, characterized in that, The weight coefficients of relevant parameters for data matching degree are determined based on the critical threshold; the weight of any relevant parameter that is greater than the critical threshold is increased, and the increase is determined based on the effective difference. Among them, the relevant parameters are the difference in historical incidence rates or the difference in rhythm disorder coefficients. The difference in historical incidence rates between the target medical record and the control medical record is recorded as the difference in historical incidence rates, and the difference in rhythm disorder coefficients between the target medical record and the control medical record is recorded as the difference in rhythm disorder coefficients.

5. The AI-based medical decision-making weight allocation method according to claim 2, characterized in that, In response to the second preset supplementary condition, the supplementary selection method is determined based on the comparison result between the correlation of effective data and the preset correlation of effective data; If the relevance of the effective data is less than or equal to the preset relevance of the effective data, then the supplementary selection method is determined to be to select medical record data based on feature keywords; If the relevance of the effective data is greater than the preset relevance of the effective data, then the supplementary selection method is determined to be to select medical record data based on the dynamic difference coefficient; The second preset supplementary condition is to determine the data extraction method as supplementing data based on effective relevance.

6. The AI-based medical decision-making weight allocation method according to claim 5, characterized in that, Determine the relevance of effective data based on keyword repetition and keyword proportion; The relevance of the effective data is positively correlated with both keyword repetition and keyword proportion.

7. The AI-based medical decision-making weight allocation method according to claim 6, characterized in that, The weighting adjustment method corresponding to the relevance of effective data is determined based on the effective coefficient of diagnosis and treatment. If the effective treatment coefficient is greater than the preset effective treatment coefficient, then the allocation will be based on the benchmark weighting method; If the effective coefficient of diagnosis and treatment is less than or equal to the preset effective coefficient of diagnosis and treatment, the weight adjustment amplitude shall be determined according to the difference in the effectiveness of diagnosis and treatment. The adjustment amplitude is positively correlated with the difference in diagnostic and treatment effectiveness.

8. The AI-based medical decision-making weight allocation method according to claim 7, characterized in that, The treatment effectiveness coefficient is determined based on the cure rate and recurrence rate. The effective treatment coefficient and the cure rate are positively correlated. The effective treatment coefficient and the recurrence rate are negatively correlated.

9. The AI-based medical decision-making weight allocation method according to claim 7, characterized in that, Data sources include medical records, laboratory data, and imaging reports.

10. The AI-based medical decision-making weight allocation method according to claim 9, characterized in that, When the weight adjustment amplitude reaches the preset adjustment amplitude, the weight adjustment ends.

Images