Related Experiment Video
Updated: Sep 12, 2025
08:17
Dynamic Lung Tumor Tracking for Stereotactic Ablative Body Radiation Therapy
Published on: June 7, 2015
15.8K
Exploration of Fully-Automated Body Composition Analysis Using Routine CT-Staging of Lung Cancer Patients for
Marc-David Künnemann1, Christian Römer1, Anne Helfen1
1Clinic for Radiology, University of Münster and University Hospital Münster, Münster, Germany.
Journal of Cachexia, Sarcopenia and Muscle
|August 6, 2025
Summary
AI-driven body composition analysis (BCA) from CT scans provides new prognostic biomarkers for lung cancer patients. The Sarcopenia Index (SI) showed the strongest survival prediction, outperforming traditional measurements.
Area of Science:
- Radiomics and Artificial Intelligence in Oncology
- Quantitative Imaging Biomarkers
- Cancer Prognostics
Background:
- Automated body composition analysis (BCA) using artificial intelligence (AI) on routine staging CT scans can yield prognostic biomarkers.
- This study investigates the prognostic value of volumetric BCA markers for overall survival in lung cancer patients across two centers.
Purpose of the Study:
- To evaluate the prognostic significance of AI-derived volumetric body composition markers in lung cancer.
- To compare the performance of these novel markers against conventional measurements for survival prediction.
Main Methods:
- Two lung cancer cohorts (Hospital A: n=3345, Hospital B: n=1364) underwent automated BCA of abdominal CT scans.
- A deep learning network segmented tissues to derive Sarcopenia Index (SI), Myosteatotic Fat Index (MFI), and Abdominal Fat Index (AFI).
- Survival analyses included Kaplan-Meier, Cox regression, and machine learning prediction, with a multivariate model validated across centers.
Main Results:
- High Sarcopenia Index (SI) predicted longer survival in nonmetastatic NSCLC for both sexes and in metastatic disease for males.
- High Myosteatotic Fat Index (MFI) was associated with reduced survival, particularly in males, with center-dependent effects in females.
- The multivariate survival model demonstrated prognostic differentiation, with SI showing significant predictive value and outperforming conventional L3 measurements.
Conclusions:
- CT-based volumetric BCA provides valuable prognostic biomarkers in lung cancer, with varying significance based on sex, stage, and center.
- The Sarcopenia Index (SI) emerged as a robust prognostic marker, surpassing traditional L3-based measurements.
- Integrating BCA markers, especially SI, into clinical workflows can enhance risk stratification and personalize lung cancer patient care.
