Collaborative Case-based Reasoning And Its Applications In Autism Diagnosis

The currently number of people with autism in China is increasing year by year,Although it has attracted widely attention,the number of professional special teachers and institutions cannot match.So it is become more and more important that how to assist normal teachers to diagnose and treat the people with autism.Since the rehabilitation training program of autistic relies heavily on the prior experience and knowledge of a professional teacher autistic,how to reuse the treating experience of specialized teachers and doctors and refer to the diagnosis and treatment plan that has been effective in the past which can to a large extent help the teachers and doctors who with poor clinical experience make new diagnosis and treatment plans.Based on this,this thesis bases on CBR(Case.Based Reasoning)presents a case-based reasoning system for diagnosis and treatment of patients with autism,assisted by doctors and special teachers in clinical intervention and diagnosis.At first,the thesis analyzed the causes of autism,Clinical manifestation and current situation and problems of diagnosis and treatment of autistic patients in China.This paper focuses on the development of rehabilitation training program and extracts the attributes of special diagnosis,the CBR is integrated into the process of the diagnosis and treatment of autism.We designed a intelligent generation system for diagnosis and treatment of autism.Secondly,according to the attributes of the diagnosis and treatment program,the knowledge base structure of the autism diagnosis and treatment system was designed.It focused on the case library of the system of diagnosis and treatment of autism.Between the design progress of case retrieval algorithm,In similarity acquisition,it is proposed to calculate the similarity based on the weighted heterogeneity matrix algorithm,in the weight optimization module the Delphi method is used to obtain weights.Meanwhile,according the attributes of the diagnosis and treatment program of autism.In case revision,a rule-based case correction method is adopted in combination with case-based case revision.At the end,the intelligent generation system of autism diagnosis and treatment was realized and the validity of the system retrieval algorithm,sensitivity and other performance indicators is verified by experiments,then compared with commonly used case-based retrieval algorithms and common classification retrieval algorithms.Through this experiments we proved the CBR algorithm has a good accuracy and comprehensive performance.In clinical practice,it has greatly helped the special teacher and the hospital doctor to make auxiliary diagnosis.