Integrated prediction method for fusing multiple major adverse cardiovascular event prediction models

Abstract

The invention relates to an integrated prediction method for fusing multiple major adverse cardiovascular event prediction models. The method comprises the following steps of binarizing output results of to-be-fused major adverse cardiovascular event prediction models, and according to the binarized output results of the prediction models, calculating a weight value of each independent prediction model by using a rough set theory; normalizing the output results of the to-be-fused major adverse cardiovascular event prediction models, and for the normalized output results of the prediction models, formatting an optimal classification threshold point to 0.5; and according to the weight values and the normalized output results, calculating basic probability assignment of each patient, combining output items of the independent prediction models by using a Dempster rule synthesis method to obtain a combined mass function, and calculating a final output value of the integrated prediction method. According to the integrated prediction method, the problem of uncertainty in models built by using electronic medical record data is solved, so that the prediction accuracy of the models is improved.

Description

technical field [0001] The present invention relates to cardiovascular disease risk assessment, in particular to an integrated prediction method that integrates multiple major adverse cardiovascular event prediction models. Background technique [0002] Acute Coronary Syndrome (ACS, Acute Coronary Syndrome), as the most serious and acute group of diseases in coronary heart disease, refers to a series of symptoms caused by the reduction of coronary blood flow, which leads to the failure of part of the myocardium to work normally or death. [0003] For the prediction of major adverse cardiovascular events in patients with acute coronary syndrome, a variety of patient characteristics can be integrated to provide a comprehensive assessment of the patient's prognosis. The evaluation results can (1) guide doctors to choose appropriate wards for patients; (2) assist doctors to rationally formulate drugs and invasive intervention treatment plans for patients, and reduce the occurren...

AI Technical Summary

Problems solved by technology

It solves the uncertainty problem of the existing adverse event prediction model based on electronic medical record data, and at the same time realizes the integration of various cohort-based research models and electronic medical record data models to build an integrated prediction model that synthesizes "expert" opinions from various aspects

Images