| China has entered the population ageing,and the incidence of dementia and cognitive impairment is increasing year by year.According to the statistical report,the elderly population will reach 300 million by 2027,the number of people with Alzheimer’s disease(AD)and mild cognitive impairment(MCI)in our country will soon exceed 60 million.Because cognitive dysfunction is a primary degenerative disease of the brain,there is no effective treatment for this disease at home and abroad,so we need to carry out early screening and diagnosis for this disease,so that doctors can early rehabilitation treatment of patients,delay the onset time and deterioration rate.At present,the severity of cognitive dysfunction in patients diagnosed in medicine often depends on the paper quality scale,according to the scale score to judge the severity level of cognitive dysfunction,but the artificial method to calculate the score is inefficient and easy to make mistakes,data storage and data mining analysis is difficult to carry out,doctors can not directly understand the development trend of patients.Therefore,the subject of this paper is to achieve accurate acquisition and mass storage of patient data through network technology,combined with traditional rehabilitation medicine,and through the design of corresponding evaluation function and intelligent evaluation model,as well as the construction of intelligent assessment model of cognitive dysfunction,it provides powerful data support for doctors’ diagnosis,and realizes the early diagnosis of AD.In this paper,we first extract the replicated intersecting pentagons from the MMSE scale as the evaluation criteria of the cognitive function intelligence assessment model,and then determine that the convolutional neural network is used to intelligently recognize the drawn images of the assessment objects,the results show that the initial recognition rate of the convolutional neural network model is 66.7%,and the model is optimized by enlarging the scale of the data and selecting the appropriate Dropout value,the recognition rate of the model can reach 89%,which shows that the convolution neural model has high accuracy and feasibility for pentagonal image recognition.Then,the paper expounds the design idea and process of the concrete function and structure of the cloud platform system and the intelligent evaluation system of cognitive function.In this paper,we choose B/S model to develop cognitive function evaluation cloud platform system and intelligent evaluation system,which can improve the universality and convenience of the system,then the specific functions and structure design of cloud platform system and intelligent evaluation system are described in detail,and the database of cloud platform system is designed in detail.Then,according to the previous function and structure design,the paper has carried on the concrete realization to the cognitive function evaluation cloud platform system and the cognitive function intelligent evaluation system,the complete operation flow of the system and the concrete use method of each function module are introduced.Then,the system is tested and the server is run and the related situation is explained.Finally,the author makes a summary of the full text,affirms the research value of cognitive function evaluation cloud platform system and intelligent evaluation system,and puts forward new prospects for future research work. |
