Research On Medical Image Classification Method Based On Deep Learning

Accurate automatic classification of medical pathology images has always been an important issue in the field of deep learning.However,the traditional manual extraction of features and image classification usually requires in-depth knowledge and more professional researchers to extract and calculate high-quality image features.This kind of operation generally takes a lot of time and the classification effect is not ideal.Aiming at the problem of medical image classification,this paper focuses on the medical image classification based on deep neural network.The main research work is as follows:In order to solve these problems,an improved Dense Net-201-MSD network model based on Dense Net-201 is proposed to implement the method for classifying medical pathology images.First,the image is preprocessed,and the traditional pooling layer is replaced by multiple scaling decomposition to prevent overfitting due to the large dimension of the image data set.Second,add BN layer,Softmax activation function and optimizer Adam in turn to improve the performance and recognition accuracy of the network model.Validate the performance of the model on the Brea KHis dataset,the new deep learning model yields image classification accuracy of 99.4%,98.8%,98.2% and 99.4% when applying to four different magnifications of pathological images,respectively.The study results demonstrate that this new classification method and deep learning model can effectively improve accuracy of pathological image classification,which indicates its potential value in future clinical application.Aiming at the problem of pneumonia image classification in medical images,the improved Dense Net-201 deep neural network model is used to classify and identify pneumonia images.Firstly,the pneumonia images of the Chest X-Ray dataset are subjected to "image inversion" processing.Secondly,a simplified multi-scale decomposition method is proposed to reduce the dimensionality of the inverted image to prevent overfitting caused by the large dimension of the image dataset.Finally,the BN algorithm,Softmax and the optimizer Adam are added to the network in turn to assist the optimization of the network model performance and classification accuracy.The experimental results show that the Dense Net-201 neural network is used to train the processed image data set significantly.