A Multi-center Gadolinium-ethoxybenzyl-diethylenetriamine Pentaacetic Acid (Gd-EOB-DTPA) MRI Dataset with Expert Annotations and clinicopathological data.
Authors
Affiliations (10)
Affiliations (10)
- Guangdong Cardiovascular Institute, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Guangzhou, 510080, China.
- Department of Radiology, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou, 510080, China.
- Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application, Guangzhou, 510080, China.
- Medical imaging center, Peking University Shenzhen Hospital, Shenzhen, 518036, China.
- Department of Radiology, The First Affiliated Hospital of Shantou University Medical College, Shantou, 515041, China.
- Department of Radiology, Shenzhen Shenshan People's Hospital, Shenzhen, China.
- Medical Research Institute, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou, 510080, China.
- Department of Radiology, The University of Hong Kong - Shenzhen Hospital, Shenzhen, China. [email protected].
- Department of Radiology, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou, 510080, China. [email protected].
- Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application, Guangzhou, 510080, China. [email protected].
Abstract
Liver resection is a cornerstone treatment for liver tumors, yet post-hepatectomy liver failure (PHLF) remains a severe and life-threatening complication with no effective treatment. Recent advances in artificial intelligence (AI) have shown promise in addressing this challenge; however, progress has been hindered by the limited availability of high-quality, expert-annotated, multi-center datasets. To bridge this critical gap, we present a multi-institutional dataset of preoperative Gd-EOB-DTPA-enhanced MRI scans, comprising 14,895 images from 220 patients across three academic medical centers. This comprehensive dataset includes 22,342 expert annotations of key anatomical structures (i.e., liver, Couinaud segments, liver tumors, spleen, and psoas muscle), detailed clinicopathological variables, and rigorously adjudicated PHLF outcomes. We also release U-Net-based automated segmentation tools to support reproducible region-of-interest delineation. Furthermore, we provide an illustrative (non-validated) pipeline demonstrating how FLR volumetry, MRI-derived functional parameters, and clinical risk factors can be integrated to support downstream analysis and methodological development. The described pipeline is illustrative and does not represent a validated predictive model. This dataset aims to facilitate research in FLR assessment and PHLF prediction, providing a resource for research on preoperative assessment.