Quantifying morphological changes in middle trapezius with ultrasound scanning and a novel histogram-matching algorithm for adults with and without Facioscapulohumeral dystrophy (FSHD)

Abstract

AbstractBackground and ObjectivesFacioscapulohumeral dystrophy (FSHD) is a neuromuscular disease causing changes in muscle structure that can negatively affect upper-limb function. Echogenicity, measured using quantified muscle ultrasound, is a potential biomarker that could be used for informing decision making. Histogram-matching allows for image normalisation, which could enable comparison of echogenicity between different machine capture settings which is a current limitation. This study aimed to investigate if ultrasonography and histogram-matching can measure trapezius muscle echogenicity and morphology for differentiating between people with and without FSHD, and different levels of arm function.MethodsSingle measurement timepoint case control study of adults with FSHD and age- and sex-matched controls. Main outcomes were trapezius echogenicity and muscle thickness measured using 2D-ultrasound, and maximum thoracohumeral elevation angle, measured using 3D-movement analysis. A sensitivity analysis evaluating the effect of histogram-matching and different reference images was conducted. Between group differences for echogenicity were evaluated using an unpaired student t-test. Echogenicity, muscle thickness and range of movement were plotted to evaluate the explained variance between variables.ResultsData was collected for 14 participants (10M:4F), seven with FSHD and seven controls with a mean (SD) age of 41.6 (15.7). Normalisation was necessary and echogenicity values for the FSHD group were higher than the controls (118.2 (34.0) vs 42.3 (14.0) respectively, with statistically significant differences (p=0.002). An overall variance of 6.2 (LLOA −2.9 to ULOA 15.4) was identified between reference images. Echogenicity accounted for the largest explained variance in muscle thickness (R2=0.81) and range of movement (R2=0.74), whilst muscle thickness and range of movement was the lowest (R2=0.61).DiscussionPeople with FSHD demonstrated higher echogenicity, smaller muscle thicknesses and less range of movement. Histogram-matching for comparison of echogenicity values is necessary and can provide quantifiable differences. Different reference images affect echogenicity values but the variability is less than between group differences. Further work is needed to evaluate the longitudinal variability associated with this method on a larger sample of people with varying levels of arm function. Ultrasound scanning and post-histogram matching may be used to quantify and compare differences in muscle structure and function people with and without FSHD.