Deep vein thrombosis detection based on deep learning for CT images

Deep vein thrombosis is one of the fatal complications of stroke and can lead to serious consequences of pulmonary embolism. It is a potentially dangerous disease that can lead to preventable mortality, which could have been avoided by timely treatment. Therefore, in this study, convolutional neural network based deep learning models were applied to detect deep vein thrombosis in computed tomography images, which is expected to be potential for aiding in early diagnosis of deep vein thrombosis as a computer-aided diagnosis system.