Test-retest reliability of 3D ultrasound to visualize the gross structures of the medial gastrocnemius

Abstract

3D ultrasound (3DUS) can be used to visualize the gross morphology of the medial gastrocnemius (MG), including muscle belly, tendon and fascicle lengths, pennation angle and muscle volume [1]. Such information can be used to indicate, and evaluate the effects of treatments that target these structures, for example in children with cerebral palsy [2]. It is essential that 3DUS is sufficiently reliable to quantify changes due to treatment at the individual level. The test-retest reliability of MG 3DUS, particularly of the fascicles, is not well established. What is the test-retest reliability of 3DUS applied on the MG of healthy adults? The MG of 16 healthy adults (27.30±6 years, 10 women, 6 men) was visualized with 3DUS with the foot in an overhanging position (Fig. 1). Two scans were carried out and participants were asked to walk approximately 50 m between scans. Muscle belly, tendon and fascicle lengths, pennation angle and muscle volume were determined from 3D reconstructions using custom-made scripts [1]. Test-retest reliability was analyzed with Bland Altmann plots to visually determine systematic differences between scans and by calculating the intraclass correlation coefficient (ICC), the relative standard error of measurement ((SEM/mean)*100%) and the relative smallest detectable difference ((SDD/mean)*100%). An intraclass correlation coefficient <0.50 was interpreted as poor, 0.50–0.75 as moderate, 0.75–0.90 as good, and >0.90 as excellent reliability [3]. No systematic differences for the morphological variables were observed between scans based on the absence of clusters in the Bland Altmann plots. ICC values were excellent (0.91-1.00) for muscle belly, tendon, and fascicle lengths, and muscle volume and good for the pennation angle (0.82). The test re-test reliability of the tendon length was found to be most reliable (ICC 1.00) with a relative SEM and SDD of 0.99% and 2.75%, respectively. Muscle belly length (%SEM 2.45%, %SDD 6.78%) and volume (%SEM 3.83%, %SDD 10.62%) were found to have better reliability than fascicle length (%SEM 5.76%, %SDD 15.97%) and pennation angle (%SEM 7.61%, %SDD 21.08%). Based on previous literature [2], the SDD values of the current study may be small enough to detect the effects of MG surgical elongation on muscle belly length and volume in children with cerebral palsy using 3DUS. However, to further elucidate the sensitivity of 3DUS, reliability and sensitivity studies should be carried out on children with cerebral palsy. Further improvements could be made to increase the accuracy of fascicle length and pennation angle determination in 3D. Fig. 1. Schematic representation of measurement set-up and analysis.Download : Download high-res image (92KB)Download : Download full-size image