Gut decisions based on the liver: prediction of colorectal neoplasia using AI-based liver analysis of routine CT scans
By
Anna Hinterberger
Jonas Bohn
Darya Trofimova
Nicolas Knabe
Julia Dettling
Tobias Norajitra
Fabian Isensee
Johannes Betge
Stefan O. Schönberg
Dominik Nörenberg
Sergio Grosu
Sonja Loges
Ralf Floca
Jakob Nikolas Kather
Klaus Maier-Hein
Freba Grawe
June 3, 2026
Clinical Scorecard: Leveraging Liver Analysis for Colorectal Neoplasia Prediction: An AI Approach Using Routine CT Scans
At a Glance
Category Detail
Condition Colorectal Neoplasia Key Mechanisms Gut-liver axis as a predictive biomarker for colorectal neoplasia. Target Population Patients undergoing colonoscopy with prior abdominal CT scans. Care Setting Clinical routine settings utilizing existing CT imaging data.
Key Highlights
AI-based liver analysis can predict colorectal neoplasia. Best-performing model achieved AUROC of 0.810 for detecting neoplasia. Sensitivity of 74.1% and specificity of 72.3% after threshold optimization. Subclassification between CRC and adenoma had AUROC of 0.674. Study supports opportunistic screening using routine CT scans. Guideline-Based Recommendations
Diagnosis
Histopathological diagnoses and colonoscopy findings define colorectal neoplasia. Management
Utilize AI-based analysis of liver features from CT scans as an adjunct to CRC screening. Monitoring & Follow-up
Monitor liver imaging features as potential biomarkers for CRC risk. Risks
Consider shared risk factors such as obesity and metabolic syndrome linking liver diseases and colorectal neoplasia. Patient & Prescribing Data
1,997 patients analyzed, including 808 with colorectal neoplasia.
AI models can enhance risk stratification for colorectal neoplasia.
Clinical Best Practices
Incorporate routine CT imaging data for opportunistic CRC screening. Apply machine learning methods for extracting hepatic features. Related Resources & Content
