SMC-LUD:Large-Scale B-Mode Liver Ultrasound Dataset for Hepatocellular Carcinoma and Hemangioma Classification.
Authors
Affiliations (13)
Affiliations (13)
- Department of MetaBioBealth, Institute for Cross-disciplinary Studies, Sungkyunkwan University, Suwon, South Korea.
- Department of Radiology, Konkuk University Medical Center, Seoul, South Korea.
- Department of Critical Care Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea.
- Department of Precision Medicine, Sungkyunkwan University School of Medicine, Suwon, South Korea.
- Department of Artificial Intelligence, Sungkyunkwan University, Suwon, South Korea.
- Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea.
- Department of Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea.
- Department of Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea. [email protected].
- Department of MetaBioBealth, Institute for Cross-disciplinary Studies, Sungkyunkwan University, Suwon, South Korea. [email protected].
- Department of Precision Medicine, Sungkyunkwan University School of Medicine, Suwon, South Korea. [email protected].
- Department of Artificial Intelligence, Sungkyunkwan University, Suwon, South Korea. [email protected].
- Personalized Cancer Immunotherapy Research Center, Sungkyunkwan University School of Medicine, Suwon, South Korea. [email protected].
- Department of Family Medicine, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, South Korea. [email protected].
Abstract
Hepatocellular carcinoma (HCC) is a leading cause of cancer-related mortality globally, and accurate classification of liver lesions using ultrasound remains challenging. We present SMC-LUD (Samsung Medical Center - Liver Ultrasound Dataset), a publicly available dataset of B-mode liver ultrasound images collected from Samsung Medical Center, Seoul, Korea, between 2015 and 2024. The dataset comprises 5,385 anonymized ultrasound images from 1,021 patients, categorized into two clinically relevant classes: hepatocellular carcinoma (images = 2,716) and hemangioma (images = 2,669). All HCC cases were histopathologically confirmed through surgical resection or biopsy, while hemangioma cases were radiologically diagnosed based on characteristic imaging features. Each image was labeled and verified by board-certified radiologists and pathologists. The dataset is organized with patient-level grouping. This resource addresses the scarcity of large, well-annotated ultrasound datasets for liver lesion classification and provides a valuable foundation for developing and validating deep learning models in liver cancer screening and diagnosis.