Multimodality annotated HCC cases with and without advanced imaging segmentation

Hepatocellular carcinoma (HCC) is the most common primary liver with incidences doubled over the past two decades due to increasing risk factors. Despite surveillance, the majority of HCC cases are diagnosed at advanced stages that can be treated only using (Transarterial chemoembolization) TACE, or systemic therapy. TACE failure can occur to 60% of patients receiving the procedure, with subsequent financial and emotional burden. Radiomics have emerged as a new tool capable of predicting tumor response to TACE from pre-procedural CT study.

This retrospectively acquired data collection includes pre- and post-procedure CT imaging studies of 105 confirmed HCC patients who underwent TACE between 2002 and 2012 with an available treatment outcome, in the form of time-to-progression and overall survival. Baseline imaging includes multiphasic contrast-enhanced CT with no image artifacts (e.g. surgical clip) and was obtained 1-12 weeks (average 3 weeks) prior to the first TACE session. Semiautomatic segmentation of liver, tumor, and blood vessels created using AMIRA was manually clinically curated. These segmentations of each pre-procedural CT study were done for the purpose of algorithm training for prediction and automatic liver tumor segmentation, and are provided here (NIfTI converted to DICOM-SEG format).

肝细胞癌（Hepatocellular carcinoma, HCC）是最常见的原发性肝脏恶性肿瘤，近二十年来，随着相关危险因素增多，其发病率已翻倍。尽管已开展筛查监测，绝大多数HCC患者确诊时已处于晚期，仅能接受经动脉化疗栓塞（Transarterial chemoembolization, TACE）或系统治疗。接受该治疗的患者中，有60%会出现TACE失败，进而带来沉重的经济与精神负担。放射组学（Radiomics）已成为一种新兴工具，可通过术前CT影像预测肿瘤对TACE的治疗应答反应。

本回顾性采集的数据集纳入了105例经确诊的HCC患者的术前及术后CT影像资料，这些患者于2002至2012年间接受了TACE治疗，且具备可用的治疗结局数据，包括进展时间与总生存期。基线影像为无图像伪影（如手术夹伪影）的多期增强CT，采集于首次TACE治疗前1至12周（平均3周）。使用AMIRA软件半自动分割得到的肝脏、肿瘤及血管区域，均经过人工临床审核校准。本次公开的所有术前CT影像分割结果均用于预测模型训练与自动肝肿瘤分割算法的开发，且已从NIfTI格式转换为DICOM-SEG格式。