CT-IDP: Segmentation-Derived Quantitative Phenotypes for Interpretable Abdominal CT Disease Classification

TL;DR

This study introduces CT-IDP, a quantitative phenotyping framework for abdominal CTs, achieving high accuracy in disease classification across multiple datasets using organ descriptors and machine learning.

Contribution

Developed and validated a comprehensive, multi-organ phenotyping framework for interpretable disease classification on abdominal CT scans, outperforming baseline models.

Findings

CT-IDP achieved macro-AUC of 0.897 on MERLIN dataset.

Outperformed baseline vision-transformer with macro-AUC of 0.880 versus 0.857 on Duke-Abdomen.

Validated externally with high performance on AMOS dataset.

Abstract

In this retrospective multi-institutional study, a quantitative phenotyping framework, CT-IDP (CT Image-Derived Phenotypes) was developed on the MERLIN abdominal CT benchmark (training, validation, and test sets- 15,175, 5,018, and 5,082 studies, respectively) and externally evaluated on two independent dataset: Duke-Abdomen (2,000) and AMOS (1,107). Multi-organ segmentations were generated with TotalSegmentator and used to derive over 900 organ and compartment-level descriptors spanning morphometry, attenuation, and contextual/burden findings. Sparse disease-specific logistic regression with elastic-net regularization was trained on MERLIN and externally validated under a frozen specification. Performance was compared against a DINOv3-based vision-transformer baseline using AUC and average precision (AP), supported by phenotype-stratified audits and coefficient-level inspection. Macro-AUC for CT-IDP versus the baseline was 0.897 versus 0.880 on MERLIN, 0.877 versus 0.857 on the Duke-Abdomen dataset, and 0.780 versus 0.756 on AMOS.