Abstract
BACKGROUND: Alterations in skeletal muscle architecture following stroke-induced neural injury serve as a major contributor to post-stroke motor impairment. Ultrasound has demonstrated promising potential for muscle assessment, however, its reliability in evaluating muscle in post-stroke patients was not systematically investigated. This systematic review aimed to synthesize available data on the reliability of ultrasound-based measurements of muscle architecture in post-stroke patients. METHODS: Literature searches were conducted across PubMed, Medline, CNKI, Embase, and Wed of Science using keywords of ultrasound (ultrasonic/echotomography) and stroke. Studies were included if they focused on the reliability of ultrasound-based muscle measurements in post-stroke patients. Data on reliability was extracted. Methodological quality was evaluated using a checklist proposed by Pretorius and Keating. The overall level of evidence was also determined. RESULTS: Of the 130,484 initial records identified, 13 studies were included. Among these, 12 were rated as high quality and 1 as low quality. Ultrasound demonstrated moderate-to-excellent reliability for measuring both muscle thickness and pennation angle in the gastrocnemius and tibialis anterior (TA) muscles, with intra-class correlation coefficient (ICC) values ranging from 0.69 to 0.99, supported by strong evidence. Furthermore, it exhibited good-to-excellent reliability in assessing both cross-sectional area (ICC = 0.76-0.97 in gastrocnemius, biceps brachii, and flexor digitorum superficialis/profundus) and muscle length (ICC = 0.89-0.98 in TA, gastrocnemius, and biceps brachii), supported by moderate evidence. Similarly, moderate evidence also supported excellent reliability of ultrasound in assessing muscle volume in the TA (ICC > 0.90). CONCLUSIONS: Ultrasound generally demonstrated moderate-to-excellent reliability for measuring muscle architecture parameters, including muscle thickness, pennation angle, cross-sectional area, muscle volume, and length, across various muscles in post-stroke patients, with supporting evidence rating from moderate to strong.