Ultrasound Segmentation of Thyroid Nodules: An Enhancement Deep Neural Network Framework

Thyroid nodules are prevalent in the human population, often identified during routine ultrasound examinations. Accurate segmentation of these nodules is crucial for diagnosing and managing potential thyroid malignancies. Nevertheless, the accurate delineation of intricate thyroid tissue poses a significant challenge attributable to the inadequate contrast and substantial noise present in ultrasound imagery. Several traditional segmentation approaches to detect thyroid nodules suffer from imprecise localization issues, while deep learning techniques offer promising solutions. This study proposes a new deep hybrid architecture that combines the encoder-decoder structure of UNet with residual blocks to enhance the model's capacity for identifying thyroid nodules. The proposed approach demonstrates the impact of introducing residual blocks to UNet architecture, and its capability to effectively capture both local and global information for accurate image segmentation because it is particularly important in thyroid nodule segmentation. The model was trained and verified using the public TN3K dataset, which consists of 3493 thyroid nodule images. Our approach was found to significantly improve the accuracy of semantic segmentation, achieving 93.56% with an IOU of 94.97% compared to state-of-the-art methods.