Agreement Between Semi-Automated CT-Based Body Composition Software and Qualitative Muscle Assessment in Sarcopenia:...

Area of Science:

• Radiology
• Medical Imaging
• Oncology

Background:

• Body composition analysis using CT scans is crucial for assessing sarcopenia in cancer patients.
• Semi-automated software tools offer potential for efficient quantitative analysis of muscle and adipose tissue.
• Qualitative muscle assessment (Goutallier classification) is commonly used but subjective.

Purpose of the Study:

• To compare quantitative body composition parameters from two semi-automated CT software tools (CoreSlicer and AsanJ-Morphometry).
• To evaluate the association between quantitative muscle/adipose tissue measures and qualitative muscle assessment (Goutallier classification).
• To assess the reliability and interchangeability of the two software tools for sarcopenia evaluation in patients with abdominal malignancies.

Main Methods:

• Retrospective study of 103 patients with abdominal malignancies undergoing CT scans.
• Independent measurement of L3 body composition parameters (TAMA, SMI, psoas area, VAT, SAT) using CoreSlicer and AsanJ-Morphometry.
• Qualitative muscle assessment via Goutallier classification; agreement assessed using ICC and Bland-Altman analysis.

Main Results:

• High agreement was found between CoreSlicer and AsanJ-Morphometry for all quantitative muscle and adipose tissue parameters.
• Higher Goutallier grades correlated significantly with lower muscle density and reduced muscle area.
• Quantitative measurements showed strong consistency with qualitative muscle assessment.

Conclusions:

• Semi-automated CT-based body composition software provides reliable quantitative data with high inter-tool agreement.
• Systematic differences exist between the software tools, limiting their interchangeability for sarcopenia assessment.
• Combining quantitative and qualitative muscle evaluation may enhance clinical sarcopenia assessment in cancer patients.