Abnormal thyroid ultrasound images

This dataset is a refined subset of the publicly available DDTI: Thyroid Ultrasound Images dataset from Kaggle, curated to focus exclusively on abnormal thyroid cases. A total of 34 grayscale ultrasound images have been selected, representing clinically relevant abnormalities such as thyroid nodules and structural irregularities. The refinement process ensures the dataset is concise, targeted, and suitable for tasks including segmentation, classification, and computer-aided diagnosis. All images have been standardized in format and anonymized for research use. This curated resource is intended to support the development and benchmarking of traditional image processing methods and advanced deep learning models in abnormal thyroid detection and segmentation.