Research On The Auxiliary Diagnosis Of COVID-19 And Alzheimer’s Disease Based On Deep Learning

Medical imaging is an important means of diagnosing diseases.Traditionally,the analysis of medical images is mainly carried out by professional physicians.However,manual analysis is often time-consuming and laborious,and requires high professional knowledge.With the development of artificial intelligence technology,computer-aided diagnosis can combine medical images with intelligent analysis algorithms to assist in diagnosis and improve diagnosis efficiency.At present,the method of combining medical imaging and deep learning has achieved good results in the auxiliary diagnosis of many diseases.This article mainly analyzes and discusses two applications:(1)Using convolutional neural networks to analyze lung CT images and classify COVID-19,other types of pneumonia,and normal lungs.(2)Using convolutional neural networks to analyze brain MRI images and classify Alzheimer’s disease,mild cognitive impairment and normal case.Coronavirus disease 2019(COVID-19)is a disease caused by severe acute respiratory syndrome coronavirus 2(SARS-Co V-2).In this paper,a convolutional neural network based on dense connection network and parallel attention module is proposed to classify COVID-19,other types of pneumonia(including bacterial pneumonia and Influenza-B virus pneumonia)and normal lung samples.Dense connection module is the basic component of feature extraction in dense connection network.The output feature maps of each convolution layer in the dense connection module are concatenated as the input feature maps of the subsequent convolution layer.This operation realizes the reuse of feature maps,which enables more features to be mined from fewer medical image samples.Besides,we use a parallel attention module that combines channel attention and spatial attention to learn the importance of features from both directions simultaneously,and assign corresponding weights according to the importance of features,so as to achieve the purpose of enhancing important features,that is,enhancing the regional features of pneumonia lesions in lung CT images.Alzheimer’s disease is a progressive neurodegenerative disease with insidious onset.Clinically,it is characterized by general dementia such as memory impairment,aphasia,apraxia,agnosia,impairment of visual spatial skills,executive dysfunction,and personality and behavior changes.Considering that the brain MRI image is three-dimensional,this paper uses a convolutional neural network using a three-dimensional convolution kernel to classify Alzheimer’s disease,mild cognitive impairment and normal samples.In addition,we design a four-dimensional attention module to learn the importance of features from four directions: channel,height,width and depth,and to strengthen the features accordingly.Experiments have proved that our network has initially achieved certain experimental results.