CN-122000015-A - Remote multidisciplinary collaborative decision-making method and system for cerebral hemorrhage emergency treatment

CN122000015ACN 122000015 ACN122000015 ACN 122000015ACN-122000015-A

Abstract

The invention discloses a remote multi-disciplinary collaborative decision-making method and system for cerebral hemorrhage emergency treatment, which are characterized in that patient image data and clinical sign data are synchronously fused to construct a dynamic diagnosis standard, on the basis, the evaluation difference of each department is analyzed, the position of opinion divergence is identified, effective consultation characteristics are extracted, target characteristic items are positioned, department diagnosis and treatment labels are defined according to the target characteristic items, decision complexity weight is generated by tracking department opinion fluctuation rules, a collaborative decision-making map is established, the collaborative decision-making map and the dynamic diagnosis standard are combined with complexity weight to complete scheme difference complementation, diagnosis and treatment conflict indexes are formed through consistency check, a dispute label table is determined, finally, an arbitration rule is established based on historical arbitration rules, weighted voting matching is completed, and a multi-disciplinary collaborative diagnosis and treatment scheme with multi-evolution path early warning is output, and quantifiable and executable multi-disciplinary collaborative decision-making support is provided for remote emergency treatment of cerebral hemorrhage emergency treatment.

Inventors

LI JIEBO
WU YANAN
WEI PENGHUI
LIN FUXIN
ZHU YANG
WANG DENGLIANG

Assignees

福建医科大学附属第一医院

Dates

Publication Date: 20260508
Application Date: 20260409

Claims (10)

1. A remote multidisciplinary collaborative decision-making method for cerebral hemorrhage emergency treatment, which is characterized by comprising the following steps: Acquiring patient image data and clinical sign data, and synchronously fusing the patient image data and the clinical sign data to construct a dynamic diagnosis benchmark; Developing condition fluctuation analysis and identification opinion bifurcation morphology through the dynamic diagnosis standard, identifying excessive consistent interval descending weight based on the opinion bifurcation morphology, retaining heterogeneous opinion to form effective consultation characteristics, and implementing condition characteristic spectrum analysis and positioning target characteristic items on the effective consultation characteristics; Defining a department diagnosis and treatment label according to the target feature item, marking a department lifting weight with high variation frequency in the department diagnosis and treatment label as a core variation department to generate a decision complexity weight, and implementing partition calibration on the dynamic diagnosis benchmark by the department diagnosis and treatment label and the decision complexity weight to establish a collaborative decision map; Performing scheme difference complementation by combining the collaborative decision-making atlas and the dynamic diagnosis benchmark with the decision complexity weight to form a correction scheme set, performing consistency check on the correction scheme set to form a diagnosis and treatment conflict index, and determining a dispute annotation table based on subject opinion deviation distribution of the diagnosis and treatment conflict index; Establishing an arbitration rule for the dispute annotation table and the department diagnosis and treatment annotation map, carrying out weighted voting matching on the dispute annotation table based on the arbitration rule to generate decision validity, and enabling the multi-evolution path early warning to be input into the output multi-disciplinary collaborative diagnosis and treatment scheme through the low consensus interval annotation of the decision validity.
2. The method of claim 1, wherein said developing a condition fluctuation resolution recognition opinion bifurcation modality by said dynamic diagnostic basis comprises: extracting disease index time sequence according to each department evaluation dimension to the dynamic diagnosis standard to generate a department index time sequence group; performing department-to-department consistency calculation on the department index time sequence group to form a consistency evaluation sequence; detecting the consistency sudden drop section for identifying a single department leading objection trigger and a multi-department synchronous bifurcation trigger of two types of sudden drops to form a sudden drop classification mark table; and respectively assigning weights to the two types of abrupt drops based on the abrupt drop classification marking table to mark the opinion disambiguation generating positions to form opinion disambiguation forms.
3. The method of claim 1, wherein labeling the department's high frequency of department's nomination in the department's diagnosis and treatment label as a core-variant department's generation decision complexity weight comprises: extracting opinion change records of each department according to time sequence to generate a change frequency sequence; performing high-frequency change department identification on the change frequency sequence to form a core change department table; performing diagnosis and treatment contribution degree lifting on the core change department table to generate core weight distribution; and generating a decision complexity weight based on the weighted integration of the core weight distribution and the variable frequency sequence.
4. The method of claim 1, wherein said complementarily forming the collaborative decision-making graph with the dynamic diagnostic benchmark in combination with the decision complexity weight completion scheme differential complement forms a set of correction schemes, comprising: performing section-by-section primary scheme matching on the collaborative decision-making map and the dynamic diagnosis reference to form a primary scheme sequence; Covering the primary scheme sequence detection scheme with a deletion segment to form a scheme deletion marker table; Extracting deletion association features of the scheme deletion marker table and the dynamic diagnosis reference to generate association deletion markers; and carrying out differential complementation by combining the decision complexity weight based on the association deletion mark to form a correction scheme set.
5. The method of claim 1, wherein the determining a dispute annotation list based on the subject opinion bias distribution of the diagnosis and treatment conflict index comprises: carrying out statistics on the opinion amplitude of each department on the diagnosis and treatment conflict index to form an opinion amplitude diagram of the department; Calculating the symmetry axis offset of the department opinion amplitude graph to form a deviation degree sequence; identifying the situation that the same department synchronously deviates from a median line in a plurality of diagnosis and treatment dimensions according to the deviation degree sequence to generate a multi-dimensional synchronous deviation mark; and extracting corresponding diagnosis and treatment conflict indexes from the department opinion amplitude graph based on the multi-dimensional synchronous deviation mark to form a dispute annotation table.
6. The method of claim 1, wherein said establishing arbitration rules for the dispute annotation tables and the department diagnosis and treatment annotation mappings comprises: Co-occurrence frequency statistics is carried out on the dispute annotation table and the department diagnosis and treatment annotation to form a co-occurrence frequency map; Extracting high co-occurrence department opinion groups and historical voting records by combining the co-occurrence frequency spectrum with the dispute annotation table to form a joint dispute mode table; Identifying opinion combinations for which the historical arbitration result is not adopted all the time in the joint dispute mode table to generate an invalid co-occurrence mark; and eliminating invalid opinion from the joint dispute mode table based on the invalid co-occurrence mark, and then carrying out merging strategy mapping to establish an arbitration rule.
7. The method of claim 1, wherein developing weighted voting match-generation decision validity for the dispute annotation table based on the arbitration rules comprises: Performing department weight area allocation on the dispute annotation table based on the arbitration rule to form a department weight distribution table; generating an internal contradiction mark by a department with the mutual contradiction of the department weight distribution table identification area and the time sequence dimension opinion; Performing weight temporary freezing on the department corresponding to the internal contradiction mark to form an effective voting weight set; and executing weighted voting matching on the dispute annotation table based on the effective voting weight group to generate decision validity.
8. A method according to claim 3, wherein said high frequency variation department identification of said variation frequency sequence to form a core variation department table comprises: extracting the variable frequency sequence according to the section-by-section variable intervals of a department to generate an interval time sequence group; continuously shrinking the interval time sequence group identification change interval to a department within a preset tolerance to form a shrinkage area mark; Performing continuous risk classification labeling on the contraction zone marks to generate a change risk mark; And generating a core change department table based on the change risk annotation.
9. The method of claim 6, wherein said identifying, for the joint dispute pattern table, combinations of opinions for which historical arbitration results have not been adopted at all times, generating an invalid co-occurrence marker comprises: extracting historic individual arbitration failure rates from the department opinions in the combined dispute mode table to form a failure rate sequence; Concentrating the failure rate sequence recognition failure on department opinion in the rapid disease progress stage to generate a progressive failure mark; Establishing a high-risk time phase arbitration file corresponding to the department opinion to generate a time phase arbitration strategy group; And generating an invalid co-occurrence mark based on the opinion combinations which are not adopted all the time in the time phase arbitration strategy group screening.
10. A remote multidisciplinary collaborative decision-making system for cerebral hemorrhage emergency treatment, comprising: the reference construction unit is used for acquiring patient image data and clinical sign data, and synchronously fusing the patient image data and the clinical sign data to construct a dynamic diagnosis reference; the feature recognition unit is used for carrying out condition fluctuation analysis and identification on the opinion bifurcation morphology according to the dynamic diagnosis standard, recognizing excessive consistent interval weight reduction and retaining heterogeneous opinions based on the opinion bifurcation morphology to form effective consultation features, and carrying out condition feature spectrum analysis and positioning target feature items on the effective consultation features; The map construction unit is used for defining a department diagnosis and treatment label according to the target feature item, marking the department lifting weight with high variation frequency in the department diagnosis and treatment label as a core variation department to generate a decision complexity weight, and implementing partition calibration on the dynamic diagnosis benchmark by the department diagnosis and treatment label and the decision complexity weight to establish a collaborative decision map; The conflict detection unit is used for completing scheme difference complementation by combining the collaborative decision-making atlas and the dynamic diagnosis benchmark with the decision complexity weight to form a correction scheme set, executing consistency check on the correction scheme set to form a diagnosis and treatment conflict index, and determining a dispute annotation table based on subject opinion deviation distribution of the diagnosis and treatment conflict index; and the arbitration output unit is used for establishing arbitration rules for the dispute annotation table and the department diagnosis and treatment annotation map, carrying out weighted voting matching on the dispute annotation table based on the arbitration rules to generate decision validity, and enabling the diagnosis and treatment proposal with multiple subjects to be input and output through the low-consensus interval annotation multiple evolution paths of the decision validity.

Description

Remote multidisciplinary collaborative decision-making method and system for cerebral hemorrhage emergency treatment Technical Field The invention relates to the technical field of emergency medical decision support, in particular to a remote multi-disciplinary collaborative decision method and system for cerebral hemorrhage emergency. Background Cerebral hemorrhage is one of critical symptoms of high mortality and high disability rate in emergency department, and has rapid disease progress and very limited treatment time window. The clinical treatment generally requires the synchronous intervention of a plurality of specialized departments such as neurosurgery, neurology, severe medical department, radiology and the like, so as to form unified diagnosis and treatment judgment. However, in the remote consultation scenario, the image data and the continuous monitoring sign data of the patient are often stored in different systems in a scattered manner, and lack of a unified time sequence integration mechanism, so that the information bases acquired by the participating departments are different, and it is difficult to develop collaborative research and judgment under the same data frame. The existing consultation flow lacks a quantitative tracking means for the opinion change rule of the department, cannot identify the fundamental opposition of the high-frequency adjustment department in the critical illness state and the interdisciplinary, and cannot perform structural evaluation on the conflict intensity when the conflict of the opinions of multiple departments occurs intensively, and also lacks an effective mediation path based on the history arbitration rule, so that the formation efficiency of a collaborative decision-making scheme is low and the degree of consensus among departments is difficult to quantify. There is therefore a need for a method to address at least one of the above problems. Disclosure of Invention The invention discloses a remote multi-disciplinary collaborative decision-making method and system for cerebral hemorrhage emergency treatment, and aims to solve the problems that multi-source heterogeneous data is difficult to integrate, the structural tracking can not be realized due to the divergence of department opinions, and a system arbitration mechanism is lacking in cross-disciplinary diagnosis and treatment conflicts in remote multi-disciplinary consultation. According to the invention, a unified time sequence fusion standard is constructed, a department opinion change rule is quantized, a conflict strength evaluation and history driving arbitration frame is established, and finally a multi-disciplinary collaborative diagnosis and treatment scheme comprising consensus main opinion and multi-evolution path early warning is generated, so that structural and quantifiable decision support is provided for remote consultation of cerebral hemorrhage emergency. The invention provides a remote multidisciplinary collaborative decision-making method for cerebral hemorrhage emergency treatment, which comprises the following steps: Acquiring patient image data and clinical sign data, and synchronously fusing the patient image data and the clinical sign data to construct a dynamic diagnosis benchmark; Developing condition fluctuation analysis and identification opinion bifurcation morphology through the dynamic diagnosis standard, identifying excessive consistent interval descending weight based on the opinion bifurcation morphology, retaining heterogeneous opinion to form effective consultation characteristics, and implementing condition characteristic spectrum analysis and positioning target characteristic items on the effective consultation characteristics; Defining a department diagnosis and treatment label according to the target feature item, marking a department lifting weight with high variation frequency in the department diagnosis and treatment label as a core variation department to generate a decision complexity weight, and implementing partition calibration on the dynamic diagnosis benchmark by the department diagnosis and treatment label and the decision complexity weight to establish a collaborative decision map; Performing scheme difference complementation by combining the collaborative decision-making atlas and the dynamic diagnosis benchmark with the decision complexity weight to form a correction scheme set, performing consistency check on the correction scheme set to form a diagnosis and treatment conflict index, and determining a dispute annotation table based on subject opinion deviation distribution of the diagnosis and treatment conflict index; Establishing an arbitration rule for the dispute annotation table and the department diagnosis and treatment annotation map, carrying out weighted voting matching on the dispute annotation table based on the arbitration rule to generate decision validity, and enabling the multi-evolution path early warning to be input into the ou