HCC-TACE-SEG-Dataset-Overview

Dataset Information

The HCC-TACE-Seg dataset is a comprehensive collection of CT images specifically designed for organ and tumor segmentation. It is part of the TCIA (The Cancer Imaging Archive) database and features data from 105 patients diagnosed with hepatocellular carcinoma (HCC). This dataset provides essential resources for predicting tumor response to transarterial chemoembolization (TACE) and improving the effectiveness of diagnostic and therapeutic strategies for HCC.

Dataset Highlights:

• Coverage: Pre- and post-treatment CT images
• Patient Range: 105 patients with HCC
• Time Span: 2002 to 2012
• Segmentation: Semi-automatically performed with manual clinical corrections
• Format: Data converted from NIfTI to DICOM-SEG for segmentation

Clinical Context: HCC is the most common primary liver cancer, and its incidence has doubled in the past two decades due to increasing risk factors. Most cases are diagnosed late, leaving TACE or systemic treatment as the primary options. Despite its prevalence, up to 60% of TACE treatments fail, underscoring the need for advanced predictive tools. The HCC-TACE-Seg dataset supports research in radiomics to predict treatment outcomes and enhance patient management strategies.

Dataset Meta Information

Dimensions	Modality	Task Type	Anatomical Structures	Anatomical Area	Number of Categories	Data Volume	File Format
3D	CT	Segmentation	Liver, Liver Tumor	Abdomen	4	628	.dcm

Resolution Details

Dataset Statistics	Spacing (mm)	Size
min	(0.58, 0.58, 2.50)	(512, 512, 40)
median	(0.78, 0.78, 2.50)	(512, 512, 87)
max	(0.98, 0.98, 5.00)	(512, 512, 185)

• Number of 2D slices in the dataset: 19,844

Label Information Statistics

Anatomy	Total Cases	Coverage	Minimum Volume (cm³)	Median Volume (cm³)	Maximum Volume (cm³)
liver	224	100%	600.28	1391.05	2804.7
liver tumor	224	100%	3.94	113.965	4054.63
portal vein	224	100%	3.29	20.3	188.73
aorta	224	100%	2.85	10.71	126.53

File Structure

In the repository, the File Structures directory contains the following files:

• structure_level_1.txt
• structure_level_2.txt
• structure_level_3.txt
• structure_level_4.txt

These files provide detailed information about the dataset structure at various levels.

Source Information

• Official Website: TCIA HCC-TACE-SEG Collection
• Download Link: HCC-TACE-SEG Dataset
• Publication Date: August 2022

Citation

Please cite the dataset using the following reference:

@article{moawad2021multimodality,
  title={Multimodality annotated HCC cases with and without advanced imaging segmentation [data set]},
  author={Moawad, A and Fuentes, D and Morshid, A and Khalaf, A and Elmohr, M and Abusaif, A and Hazle, JD and Kaseb, AO and Hassan, M and Mahvash, A and others},
  journal={The Cancer Imaging Archive},
  year={2021}
}

---