Study investigates the performance of a deep learning-based model for lung segmentation from CT images for normaland COVID-19 patients . Chest CT images and corresponding lung masks of 1200 confirmed COV-19 cases were used for training a residual neural network . The proposed deep learning methodachieved DSC of 0.980 and 0.971 for normal and COVI-19 subjects, respectively,demonstrating significant overlap between predicted and reference lung masks . The promising results achieved by the proposed deeplearning-based . model demonstrated its reliability in . COVE-19 lungsegmentation . This prerequisite step would lead to a more efficient and robust . robustpneumonia lesion analysis. This prerequisite . analysis would . lead to . a more . robust and robust and . robustPneumonia . analysis . of the lung tissue analysis. of the normal and CoV-induced infections. The model was found to have a similar performance inlung segmentation of the . normal patients, respectively. was observed to the normal patients. The comparable performance of the model indicates the accuracy of . the model, though slightly better performance was observed for normal patients’s lung tissue

Author(s) : Faeze Gholamiankhah, Samaneh Mostafapour, Nouraddin Abdi Goushbolagh, Seyedjafar Shojaerazavi, Parvaneh Layegh, Seyyed Mohammad Tabatabaei, Hossein Arabi

Links : PDF - Abstract

Code :
Coursera

Keywords : lung - normal - model - patients - analysis -

Leave a Reply

Your email address will not be published. Required fields are marked *