A novel computed tomography method to detect normal from abnormal psoas muscle: a pilot feasibility study

Abstract

Background

Sarcopenia is a syndrome characterized by progressive loss of skeletal muscle which can be detected by computed tomography (CT) by estimating total psoas muscle cross-sectional area (CSA). Relying on total psoas CSA alone takes into account abnormal muscle and intramuscular fat, both of which may be increased in sarcopenic obesity. We developed a novel CT-method to identify the proportion of normal to abnormal psoas muscle at the third lumbar (L3) level. The primary objective of our pilot study was to measure inter-observer agreement between measuring total psoas CSA and proportion of normal and abnormal psoas muscle using a novel CT-method. We hypothesized total psoas CSA and proportion of normal and abnormal psoas muscle would be reliably quantifiable.

Methods

CT abdomen images were obtained for 20 adults. Two radiologists independently identified and traced the L3 psoas muscle circumference to estimate CSA. Hounsfield units were applied to the tracing to identify proportion of normal muscle, abnormal muscle, and fat. Inter-observer agreement was assessed using Pearson's correlation coefficient.

Results

Of the 20 patients, 13 were male and six were obese. Mean age was 66 years. Correlation coefficient was excellent for total psoas CSA (r=0.93, p-value<0.00001) and proportion of normal psoas muscle (r=0.94, p-value<0.0001). Correlation was excellent between BMI and abnormal muscle (r=0.67, p-value=0.001). Correlation was poor between total psoas CSA and body mass index (BMI) (r=0.369, p-value=0.108) and negative between proportion of normal muscle and BMI (r= -0.50, p-value=0.025).

Conclusions

Our study findings demonstrate that total psoas CSA and proportion of normal and abnormal psoas can be reliably quantified. Our CT-method may be superior to total psoas CSA in identifying sarcopenic obesity, the results of which can be used to explore clinical outcomes.