LungNet:Integrating CNN with channel attention and multi-scale transformer

【摘要】 The SARS-CoV-2 virus has caused various health problems worldwide, including coughing and wheezing. Computed tomography(CT) imaging of lungs can help to determine the presence and location of disease. However, manually evaluating large numbers of CT images by healthcare professionals places strict demands on their expertise. Our team developed a LungNet system to analyze CT images with the goal of detecting the presence of disease, characterizing the type of lesion to aid medical professionals in diagnosis. To evaluate the performance of our model, we conduct experiments on the publicly available SARS-CoV-2 CT scan dataset, and the classification accuracy can reach 98.8%.更多还原

【Abstract】 The SARS-CoV-2 virus has caused various health problems worldwide, including coughing and wheezing. Computed tomography(CT) imaging of lungs can help to determine the presence and location of disease. However, manually evaluating large numbers of CT images by healthcare professionals places strict demands on their expertise. Our team developed a LungNet system to analyze CT images with the goal of detecting the presence of disease, characterizing the type of lesion to aid medical professionals in diagnosis. To evaluate the performance of our model, we conduct experiments on the publicly available SARS-CoV-2 CT scan dataset, and the classification accuracy can reach 98.8%.更多还原