摘要：As the most common malignancy in the endocrine system, thyroid cancer is usually diagnosedby discriminating the malignant nodules from the benign ones using ultrasonography, whose interpretationresults primarily depends on the subjctivity judgement of the radiologists. In this study, we propose a novelcascade deep learning model to achieve automatic objective diagnose during ultrasound examination forassisting radiologists in recognizing benign and malignant thyroid nodules. First, the simplified U-net isemployed to segment the region of interesting (ROI) of the thyroid nodules in each frame of the ultrasoundimage automatically. Then, to alleviate the limitation that medical training data are relatively small in size,the improved Conditional Variational Auto-Encoder (CVAE) learning the probability distribution of ROIimages is tained to generate new images for data augmentation. Finally, ResNet50 is trained with bothoriginal and generated ROI images. As consequence, the deep learming model formed by the trained U-netand trained Resnet-50 cascade can achieve malignant thyroid nodule recognition with the accuracy of87.4%, the sensitivity of 92%, and the specificity of 86.8%.
其他摘要：As the most common malignancy in the endocrine system, thyroid cancer is usually diagnosed by discriminating the malignant nodules from the benign ones using ultrasonography, whose interpretation results primarily depends on the subjectivity judgement of the radiologists. In this study, we propose a novel cascade deep learning model to achieve automatic objective diagnose during ultrasound examination for assisting radiologists in recognizing benign and malignant thyroid nodules. First, the simplified U-net is employed to segment the region of interesting (ROI) of the thyroid nodules in each frame of the ultrasound image automatically. Then, to alleviate the limitation that medical training data are relatively small in size, the improved Conditional Variational Auto-Encoder (CVAE) learning the probability distribution of ROI images is trained to generate new images for data augmentation. Finally, ResNet50 is trained with both original and generated ROI images. As consequence, the deep learning model formed by the trained U-net and trained Resnet-50 cascade can achieve malignant thyroid nodule recognition with the accuracy of 87.4%, the sensitivity of 92%, and the specificity of 86.8%.
