Primary Liver Cancer CECT Imaging Dataset

Primary liver cancer is a significant global health issue, characterized by high incidence and mortality rates worldwide. Accurate diagnosis and classification of subtypes are essential for selecting appropriate treatment options and enhancing patient outcomes. Contrast-enhanced computed tomography (CECT) has proven highly sensitive and specific in diagnosing liver cancer. Currently, publicly available datasets of liver cancer CECT scans are limited and often do not comprehensively cover liver cancer subtypes or include complete phasing of CT scans. We hypothesize that utilizing full-phase 3D CECT images, including the Plain, Arterial, Venous, and Delayed phases, can improve the diagnostic classification performance for liver cancer. To test this hypothesis, we have collected a large dataset from a single medical institution that includes 278 cases of liver cancer, featuring Hepatocellular Carcinoma (HCC), Intrahepatic Cholangiocarcinoma (ICC), and Combined Hepatocellular-Cholangiocarcinoma (cHCC-CCA), as well as CECT images from 83 non-liver cancer subjects. For each patient, we annotated the liver and lesion regions. This dataset, rich in liver cancer types and complete in CT phasing, facilitates the development and validation of diagnostic classification models and lesion segmentation models tailored to liver cancer CT imaging.The median age of participants was 59 years [51, 67] (IQR), with 185 males (67.3% of the liver cancer group) . Each patient had complete 3D contrast-enhanced CT (CECT) data across the Plain, Arterial, Venous, and Delayed phases, stored as NIFTI files. A total of 50,560 slices containing lesions were collected, with a median lesion volume of 75.37 cm³ [26.70, 239.24] . The Python code for loading and processing the data can be found on GitHub (https://github.com/ljwa2323/PLC_CECT).原发性肝癌是一项严峻的全球公共卫生问题，在全球范围内兼具高发病率与高死亡率特征。精准诊断与亚型分型对于选择适宜治疗方案、改善患者预后至关重要。增强计算机断层扫描（Contrast-enhanced computed tomography, CECT）已被证实对肝癌诊断具备极高的灵敏度与特异度。当前，公开可获取的肝癌CECT影像数据集较为匮乏，且往往未能全面覆盖肝癌亚型，亦未包含完整时相的CT扫描影像。我们提出假设：使用包含平扫期、动脉期、静脉期及延迟期在内的全时相三维CECT影像，可提升肝癌诊断分型的性能。为验证该假设，我们从单一医疗机构采集了大型数据集：其中包含278例肝癌病例，涵盖肝细胞癌（Hepatocellular Carcinoma, HCC）、肝内胆管癌（Intrahepatic Cholangiocarcinoma, ICC）以及肝细胞癌-肝内胆管细胞癌混合型（Combined Hepatocellular-Cholangiocarcinoma, cHCC-CCA），同时纳入83例非肝癌受试者的CECT影像。针对每一例受试者，我们均对肝脏及病灶区域进行了标注。本数据集肝癌亚型丰富且CT时相完整，可为适配肝癌CT影像的诊断分型模型与病灶分割模型的开发与验证提供支撑。受试者的年龄中位数为59岁[四分位间距：51, 67]，其中肝癌组男性共185例，占该组总人数的67.3%。每例受试者均拥有覆盖平扫、动脉、静脉及延迟期的完整三维增强CT（CECT）数据，以NIFTI格式存储。本次共收集到50560张含病灶的CT切片，病灶体积的中位数为75.37 cm³[四分位间距：26.70, 239.24]。本数据集的数据加载与处理Python代码可于GitHub平台获取，链接为：https://github.com/ljwa2323/PLC_CECT。

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于原发性肝癌的医学影像资源，包含278例肝癌患者（涵盖肝细胞癌、肝内胆管癌和混合型肝癌三种亚型）和83例非肝癌对照的完整四相位（平扫、动脉期、静脉期、延迟期）3D CECT扫描图像，所有数据均以NIFTI格式存储并标注了肝脏和病灶区域。其特点是覆盖全面的肝癌亚型和完整CT相位，提供了50,560个病灶切片，旨在支持肝癌诊断分类和病灶分割算法的开发与验证，弥补了公开数据中肝癌亚型覆盖不足和相位缺失的局限。