Reliability of a novel ultrasound system for gray-scale analysis of muscle

Introduction: Ultrasound is increasingly used as an adjunct in the diagnosis of neuromuscular disease by measuring muscle thickness and echointensity (EI). Reproducibility is limited because of variations in scanning technique and proprietary algorithms that alter EI values. Methods: We developed a standardized scanning protocol and a portable machine without any postimaging processing. Ten subjects underwent scanning of 6 muscles by 3 sonographers on 2 separate days. One of the sonographers repeated the protocol with 4 different machine/transducer combinations. Gray-scale values were measured from each image with the use of a region of interest (ROI) box. Results: Combined intraclass correlation coefficients were 0.92 (intra-rater), 0.88 (inter-rater), and 0.96 (inter-system). The biceps had the highest variability (coefficient of variance [COV] 12.7%), and the medial gastrocnemius had the lowest variability (COV 7.4%). Conclusions: We demonstrate excellent reliability of a reproducible ultrasound system for gray-scale analysis of muscle that has potential applicability as a screening tool for neuromuscular disease. Muscle Nerve 56: 408–412, 2017.