Ultrasound Signal Processing Technique for Subcutaneous-Fat and Muscle Thicknesses Measurements

Accurate body-composition measurements are important for diagnosing health status. Devices used for body-composition measurement should be easily accessible for patient diagnosis whilst realizing high inter- and intra-operator accuracies for repetitive measurements. Dual-energy X-ray absorptiometry and magnetic resonance imaging (MRI) have been considered as a gold-standard method. However, these methods have disadvantages such as limited accessibility, high costs, and long scanning times. Ultrasound imaging is an alternative technique for body-composition measurement owing to its easy accessibility and convenience of use. Current ultrasound imaging techniques identify the interface between different tissue layers based on echogenicity changes observed in ultrasound images. Their measurement accuracies, therefore, depend on the ultrasound image quality and operator interpretation. Radio-frequency (RF) signals obtained directly from an ultrasound system ensure the reproducibility of measurements. However, RF signals contain substantial noise, and signal processing is fundamental in body-composition measurements. This study proposes an ultrasound signal processing technique to measure body composition. Backscattered RF signals in ex vivo swine-tissue samples were first acquired from an ultrasound system. Subsequently, interfaces of subcutaneous-fat, muscle, and bone were identified during RF signal processing. Next, subcutaneous-fat and muscle thicknesses were calculated based on the speed of sound through tissue. Lastly, the subcutaneous-fat and muscle thicknesses measured using ultrasound signals were validated using MRI scans. A strong linear correlation was observed between the proposed ultrasound method and MRI. Thickness correlations between ultrasound and MRI were observed to be 0.899 and 0.982 for subcutaneous-fat and muscle, respectively. The proposed technique, therefore, demonstrates clinical potential for body-composition measurements.