Colorectal Liver Metastases: CT Scans and Segmentations from 197 Patients

197 patients' preoperative CT scans and manual radiologist segmentations from Memorial Sloan Kettering Cancer Center. The dataset contains portal-venous phase contrast-enhanced MDCT scans of the liver, released under TCIA in 2023. It was described in a Scientific Data publication by Simpson et al. in 2024.

Use Cases

• Training segmentation models for liver metastases based on manual radiologist annotations.
• Developing computer-aided diagnosis tools for colorectal cancer based on contrast-enhanced CT scans.
• Benchmarking medical image analysis algorithms on a clinically curated dataset.
• Studying the morphology and distribution of liver metastases in preoperative imaging.

Strengths

• Includes manual segmentations by radiologists, providing high-quality ground truth labels.
• Contains data from 197 patients, offering a substantial sample size for medical imaging research.
• Scans are from a single, reputable institution (Memorial Sloan Kettering Cancer Center), ensuring consistent acquisition protocols.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect the specific patient population and imaging protocols of a single cancer center.

Provenance

Source

Memorial Sloan Kettering Cancer Center (MSKCC), released under TCIA.

Collection Method

Preoperative contrast-enhanced CT scans collected from patients undergoing hepatic resection.

Freshness

Last updated 2026-05-21 15:48:17; freshness should be verified.