Abstract
This study aimed to determine the within- and between-session reliability of medial gastrocnemius (MG) architecture (e.g. muscle thickness (MT), fascicle length (FL) and pennation angle (PA)), as derived via ultrasonography followed by manual digitization. A single rater recorded three ultrasound images of the relaxed MG muscle belly for both legs of 16 resistance trained males, who were positioned in a pronated position with their knees fully extended and the ankles in a neutral (e.g. 90°) position. A subset of participants (n = 11) were retested under the same conditions ~48-72 hours after baseline testing. The same rater manually digitized each ultrasound image on three occasions to determine MG MT, FL and PA before pooling the data accordingly to allow for within-image (n = 96), between-image (n = 32) and between-session reliability (n = 22) to be determined. Intraclass correlation coefficients (ICCs) demonstrated excellent within-image (ICCs = 0.99-1.00, P < 0.001) and very good between-image (ICCs = 0.83-0.95, P < 0.001) and between-session (ICCs = 0.89-0.95, P < 0.001) reliability for MT, FL and PA. Between-session coefficient of variation was low (≤ 3.6%) for each architectural parameter and smallest detectible difference values of 10.6%, 11.4% and 9.8% were attained for MT, FL and PA, respectively. Manually digitizing ultrasound images of the MG muscle at rest yields highly reliable measurements of its architectural properties. Furthermore, changes in MG MT, FL and PA of ≥ 10.6%, 11.4% and 9.8% respectively, as brought about by any form of intervention, should be considered meaningful.