In this project, a classification model has been made for the detection and classification of Thyroid Nodules from ultrasound images. They are categorized using Thyroid Imaging Reporting and Data System (TI-RADS).
The dataset used is from the Digital Database of Thyroid Ultrasound Images (DDTI), which contains around 400 ultrasound images of thyroid nodules. The dataset is divided into JPG files and XML files. Each image in these JPG files has a corresponding XML file and each image has a resolution of 560x360. These XML files provide a detailed classification for each of the nodules. The categories of the TIRADS in this dataset are “2” (Benign), “3” (Malign), “4a” (Malign), “4b” (Malign), “4c” (Malign), and “5” (Malign).