Research On Intelligent Diagnosis Model Based On Machine Learning

At this stage,due to the uneven distribution of medical resources in my country,high-quality medical resources are mostly concentrated in large hospitals,so patients will prefer to choose large hospitals when they go to the hospital,which causes a high number of outpatients in large hospitals.Moreover,most patients have limited medical knowledge,and most of them are unable to choose the right department and suitable doctor according to their own conditions.Therefore,the current hospital halls are equipped with manual consultation tables to help patients choose the appropriate hospital department or doctor.The traditional manual guidance method makes it difficult to make accurate department division and doctor recommendation for patients in a short period of time when facing a large number of patients,resulting in an unsatisfactory patient consultation process and reduced patient experience.At present,the hospital has opened the functions of appointment registration and online manual consultation,which has relieved the work pressure of offline consultation staff to a certain extent.However,merely shifting the offline manual consultation work to online not only increases the human and financial resources invested by the hospital,but also makes it difficult to fundamentally improve the patient registration efficiency and medical experience.How to quickly and accurately guide patients to register and select doctors for treatment is a problem that needs to be solved urgently in the guidance work of large-scale comprehensive hospitals at this stage.This paper aims at the problem that traditional manual guidance is difficult to meet the needs of patients in large hospitals,and proposes an intelligent guidance model based on machine learning.The main research object of the model is to quickly and accurately visit patients in large hospitals,and use machine learning as a research method to solve the two key problems faced by large hospitals in the guidance work-rapid and accurate department division and physician recommendation.This article analyzes the functions of the intelligent guidance model and divides the intelligent guidance model proposed in this article into three parts.The first is the research of symptom entity extraction method of patient's complaint information.The traditional guidance model mostly adopts the way of patient's independent selection to collect the patient's symptoms,but the patient's own medical knowledge level is limited,which leads to inaccurate selection of symptoms and affects the accuracy of the guidance.Based on this problem,this paper proposes a symptom entity extraction method based on the combination of word features the symptoms of the patients were extracted from the chief complaint information of the patients,and the effectiveness of the method was proved by multiple control experiments;Secondly,this paper takes the triage problem as a multi label classification problem,and proposes a stacking method of department triage based on symptom contribution and information entropy weighting.The experimental results show that the method can complete the task of triage,and compared with the traditional classification methods,the accuracy of the results is significantly improved;Thirdly,based on the results of triage,this paper proposes a physician recommendation model based on patient similarity and physician matching.It mainly introduces patient similarity and physician matching degree to narrow the range of adjacent patients,so as to improve the accuracy and efficiency of physician recommendation The results show that the performance of this method is better than the traditional recommendation method in accuracy,recall rate and time consumption.