Research On Medical Image Classification Based On Multi-Layer Network

In recent years,medical image automated analysis has become an important research topic in the field of medical imaging.While there are different imaging techniques such as diagnostic X-rays,magnetic resonance imaging and ultrasonography to detect and diagnose tumors,the analysis of diseased tissue by an experienced specialist is the only way to diagnose tumors with confidence.Medical image analysis is a very time-consuming and specialist task,depending on the skills and experience of the specialist.After all,diagnosis is subjective and can be affected by several factors such as fatigue and distraction.In addition,there is often a lack of consensus among experts in terms of diagnosis.The continuous accumulation of clinical cases has led to an increased demand for medical image analysis.Based on this fact,there is an urgent need for computer-aided diagnostic systems capable of automatic detection and classification to reduce the workload of professionals and increase efficiency.Aiming at the problem of biomedical image automatic classification based on tensor networks,an Unordered Stacked Tensor Network model(UnSTNet)is proposed.Tensor networks have been an enabling tool for the analysis of quantum many-body systems in physics in the past and now are applied to medical image analysis tasks.We extend the matrix product state to adapt to the medical image analysis task,and embed the classical image domain concept,local disorder of images,to preserve the global structure of images.In addition,we stack unordered blocks within tensor blocks to integrate global information and stack the outputs of multiple tensor blocks to fuse image features in different states for global evaluation.We evaluate on three publicly available histopathology image datasets and demonstrate that the proposed method obtains improved performance compared to related tensor learning methods.The results also show that,compared with the advanced deep learning method,our model can perform well with fewer computing resources and hyperparameters.Aiming at the problem of biomedical image automatic classification based on deep learning networks,a Dual-ended Multiple Attention Learning model(DMAL)is proposed.The model incorporates multiple attention learning into both networks,and the two networks are linked using an integration module.Specifically,in both networks,the backbone network is used to extract global features and the branch network captures local area information;the integration module combines multi-stage features;and the attention module containing element,channel and spatial attention prompts the model to focus on multi-scale information relevant to the disease.We conducted experiments on the COVID-19 image dataset.Meanwhile,we evaluate the proposed DMAL network using relevant competitive methods as well as ten advanced deep learning models in the image domain and obtain the best performance with 99.67%,99.53%,99.66%,99.60% and 99.76% in terms of Accuracy,Precision,Sensitivity,F1 Scores and Specificity.The proposed method will help in the rapid screening and high-precision diagnosis of COVID-19,given the general trend of such severe global infections.