Answers are grounded in this paper and cite their sources.
Narrative review highlights discrepancies in CT body composition methods and emerging AI technologies for clinical outcomes.
Body composition research utilizing computed tomography (CT) has increased over the past several decades as researchers use clinically acquired CT for opportunistic screening or the identification of phenotypes such as sarcopenia, myosteatosis, and sarcopenic obesity associated with various clinical outcomes. While there continues to be exciting work being done in CT body composition research, including the use of artificial intelligence (AI) to streamline the assessment of body composition, fundamental methodological discrepancies continue to hinder clinical integration. In this narrative review of studies published between 2022 and 2025, we describe newer approaches to normalize skeletal muscle measurements for differences in body size. We provide an overview of the heterogeneity in deriving currently published CT reference criteria for body composition quantity and quality in a limited subset of studies from published literature. We also summarize currently published thresholds from a limited subset of studies, including those most commonly cited, applied to define low muscle quantity or poor muscle composition and compare and contrast these values, as well as their derivation. Finally, we highlight emerging areas utilizing AI technologies and the integration of CT body composition assessment into clinical practice, and on the flip side, the lack of guidance on how this information will be implemented to achieve personalized nutrition care and improve clinical outcomes.
No takes yet. Share an insight, caveat, or question.
Byrne et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: