CONSTRUCT AN APPROXIMATION DECISION MODEL OF MEDICAL RECORD BY NEURAL NETWORKS FOR CLINICAL DIAGNOSIS GUIDELINE — THE OPHTHALMOLOGY DEPARTMENT AS AN EXAMPLE

A doctor often has to independently make decisions for the therapy or treatment of patients, such as during clinical emergency calls, some of the most challenging tasks for a doctor. Only if a doctor has highly sophisticated "professional knowledge" and rich "diagnosis and treatment experiences", can he or she be competent for these tasks. Therefore, the inheritance of experience from clinical emergency calls is an important connection in the formative education of doctors. The most traditional formative education of doctors is to have students simulate the diagnosis and treatment modes of their teachers, or carry handbooks containing clinical summaries to perform a diagnosis with them as guidelines. However, this temporary approach of finding answers in books is not only time-consuming, but also detrimental to the confidence of the patients in their doctors. Our research involves the aforementioned clinical practices: clinical emergency calls and doctor formative education. We use ophthalmology as an example and utilize the back-propagation algorithm of the artificial neural networks, to construct an approximate decision model of medical record to guide clinical diagnosis. The doctor can input information such as chief complaint, minor complaints, diagnosis, etc. into the decision model. Correct ophthalmologic approximation medical records are created by referencing diagnosis and treatment to improve the quality of medical treatment and medical care.