AI-based volumetric six-tissue body composition quantification from CT cardiac attenuation scans for mortality...

Area of Science:

• Radiology
• Artificial Intelligence
• Cardiovascular Imaging

Background:

• CT attenuation correction (CTAC) scans are standard in cardiac perfusion imaging but underutilized.
• Current uses include attenuation correction and visual calcium scoring.
• Novel AI approaches can extract more diagnostic information from these scans.

Purpose of the Study:

• To develop an AI-based method for volumetric body composition analysis from CTAC scans.
• To assess the prognostic value of these body composition measures for all-cause mortality.

Main Methods:

• AI-based segmentation and image processing were applied to CTAC scans from a multi-site international registry.
• Volumetric measurements of bone, skeletal muscle, and adipose tissues (subcutaneous, visceral, intramuscular, epicardial) were quantified.
• Cox regression and Kaplan-Meier analyses evaluated the association between body composition measures and all-cause mortality.

Main Results:

• AI processing took less than 2 minutes per scan.
• In 9918 patients, high visceral adipose tissue (VAT), epicardial adipose tissue (EAT), and intramuscular adipose tissue (IMAT) were linked to increased mortality risk.
• High bone attenuation and skeletal muscle volume index were associated with reduced mortality risk.

Conclusions:

• CTAC scans contain valuable, automatically measurable body composition biomarkers.
• These AI-derived measures provide significant additional prognostic information for all-cause mortality.
• This approach enhances the utility of routine cardiac imaging scans.