Diagnosis method of auxiliary diagnosis model based on disease binary classifier

Abstract

The invention discloses a diagnosis method of an auxiliary diagnosis model based on a disease binary classifier. The diagnosis method comprises the following steps: S1, data preprocessing; S2, model training; S3, diagnosing and predicting. According to the method, diagnosis prediction is defined as a text classification task, namely, for an input patient EHR, information such as chief complaint, current medical history and iconography is extracted, whether a patient suffers from diseases or not is predicted by training a binary classifier (BiLSTM + Self-Attention model) of multiple diseases, and finally prediction of patient diagnosis is obtained so as to assist a doctor in making a later decision.

Description

technical field [0001] The invention belongs to the field of medical technology, and in particular relates to a diagnosis method based on an auxiliary diagnosis model of a disease binary classifier. Background technique [0002] With the development of medical informatization, the number and scale of electronic medical records (EHR) have been increasing, forming a huge electronic database that integrates a variety of clinical information. Therefore, using artificial intelligence methods to mine information in EHR data has become a potentially powerful tool for disease diagnosis and management. As a tool to assist doctors in clinical decision-making, assisted diagnosis uses machine learning technology to extract relevant clinical information of patients (main complaint, history of present illness, imaging, etc.) from EHR text, and simulates the clinical reasoning of doctors to accurately predict the diagnosis of patients. [0003] Auxiliary diagnosis can be regarded as a tas...

AI Technical Summary

Problems solved by technology

However, the multi-label classification model has the problem of uncertain number of class labels and a...

Images