A preclinical CT and MRI Liver Imaging Dataset with Anatomical, Functional and Segmentation Data

TL;DR

This paper introduces a new preclinical liver imaging dataset with MRI and CT scans from mice, annotated for use in research on liver diseases.

Contribution

The first publicly available preclinical liver imaging dataset combining anatomical, functional, and segmentation data from multiple mouse models.

Findings

The dataset includes MRI and CT scans from mice with hepatocellular carcinoma, MASH, and fibrosis.

Some scans have annotated segmentations, enabling training of automated image analysis tools.

Metadata is structured using a tailored ISA-Tab profile for improved data reuse and reproducibility.

Abstract

Chronic liver diseases (CLD) account for more than 2% of deaths worldwide. Extensive research has been conducted to better understand CLD, generating vast amounts of data. However, only a small fraction of raw preclinical data are publicly available, posing a significant challenge for transparency, reproducibility, and data reuse. Therefore, we built a preclinical liver imaging dataset, the first of its kind to our knowledge. The database contains longitudinal liver MRI scans from mice with hepatocellular carcinoma, metabolic dysfunction-associated steatohepatitis (MASH, formerly NASH), and fibrosis, as well as CT scans of mice with MASH and mice carrying a dysfunctional ICAM-1 gene. Superimposable MRI and CT scans bridge the gap between the modalities. Some of the 222 murine scans have annotated segmentations. Metadata containing both scan and mouse parameters are organized using a tailored metadata profile in ISA-Tabs. This dataset enables advanced image analysis, such as building tools for automated segmentation, train radiomics analysis tools, or can be used as a reference control dataset.

Figures