Related Experiment Video
Updated: Mar 13, 2026
07:03
Imaging and Quantification of the Hepatic Vasculature of Mice Using Ultrafast Doppler Ultrasound
Published on: July 19, 2024
1.9K
SMC-LUD:Large-Scale B-Mode Liver Ultrasound Dataset for Hepatocellular Carcinoma and Hemangioma Classification
Jisoo Tak1,2, Ryoung-Eun Ko3, Ryan Donghan Kwon4
1Department of MetaBioBealth, Institute for Cross-disciplinary Studies, Sungkyunkwan University, Suwon, South Korea.
Scientific Data
|March 12, 2026
Summary
A new dataset of 5,385 liver ultrasound images aids hepatocellular carcinoma (HCC) classification. This resource supports AI development for liver cancer diagnosis.
Area of Science:
- Medical Imaging
- Oncology
- Data Science
Background:
- Hepatocellular carcinoma (HCC) is a major global cause of cancer mortality.
- Accurate classification of liver lesions via ultrasound is difficult.
- There is a lack of large, well-annotated ultrasound datasets for liver lesion classification.
Purpose of the Study:
- To introduce the Samsung Medical Center - Liver Ultrasound Dataset (SMC-LUD).
- To provide a valuable resource for developing and validating AI models for liver cancer diagnosis.
Main Methods:
- Collected 5,385 B-mode liver ultrasound images from 1,021 patients (2015-2024).
- Categorized images into hepatocellular carcinoma (2,716) and hemangioma (2,669).
- Ensured diagnostic accuracy through histopathological confirmation (HCC) and radiological diagnosis (hemangioma), with radiologist and pathologist verification.
Main Results:
- Established the SMC-LUD, a publicly available dataset.
- Dataset includes 5,385 anonymized images with patient-level organization.
- Contains histopathologically confirmed HCC and radiologically diagnosed hemangioma cases.
Conclusions:
- The SMC-LUD addresses the scarcity of annotated liver ultrasound data.
- This dataset will facilitate the advancement of deep learning models for liver cancer screening and diagnosis.
