RESUMO
Introduction Currently, little attention is given to the muscle architecture reliability studies of the hamstring using a robust statistical. Our purpose was to determine the reliability of ultrasound measurements of muscle thickness, fascicle length and pennation angle of the biceps femoris and semitendinosus muscles, including heteroskedasticity and internal consistency analyses. Methods Two images of biceps femoris and semitendinosus at 50% of the thigh length were acquired from 21 volunteers, in two visits. The parameters were measured three times in each image, and for each muscle. The reliability was analyzed by the intraclass correlation coefficient (ICC) and Cronbach’s alpha (αCronbach). The relative standard error of the measurements (%SEM) were calculated and Bland-Altman plots were generated. Results All parameters presented excellent ICC for the three repeated measurements (ICC from 0.93 ‒ 0.99) and moderate to excellent reliability intraday (ICC from 0.70 ‒ 0.95) for both muscles. The present study indicates that ultrasound is a reliable tool to estimate the biceps femoris fascicle length (ICC = 0.97, αCronbach = 0.98, %SEM = 7.86) and semitendinosus (ICC = 0.90, αCronbach = 0.95, %SEM = 7.55), as well as the biceps femoris muscle thickness (ICC = 0.89, αCronbach = 0.94, %SEM = 10.23) and semitendinosus muscle thickness (ICC = 0.87, αCronbach = 0.93, %SEM = 1.35). At last, biceps femoris pennation angle (ICC = 0.93, αCronbach = 0.96 and %SEM = 4.36) and semitendinosus (ICC = 0.96, αCronbach = 0.98 and %SEM = 4.25) also had good repeatability. Conclusion Ultrasonography show good repeatability in estimating of muscle architecture parameters.