Quantitative ultrasonography of the foot muscles: a comprehensive perspective on reliability.

Abstract

Background: Quantitative ultrasound imaging is a popular technique to assess the structural properties of the intrinsic and extrinsic foot muscles. Although several studies examined test-retest reliability, specific gaps remain in assessing inter-rater reliability, particularly distinguishing between image acquisition and muscle measurement. Additionally, these studies utilized equipment that may not be generalizable across both clinical and research settings and often involved small sample sizes without prior sample size calculations. This study aimed to investigate test-retest reliability as well as global and measurement-based inter-rater reliability (MIRR) using a low-end ultrasound device to measure intrinsic and extrinsic foot muscle sizes.

Methods: This prospective reliability study included 21 active individuals. Five intrinsic muscles [abductor hallucis (AbH), flexor digitorum brevis (FDB), flexor hallucis brevis (FHB), quadratus plantae (QP), abductor digiti minimi (AbDM)], and three extrinsic muscles [peroneal (PER), flexor digitorum longus, tibialis anterior (TA)] were scanned. Three investigators independently acquired images on two occasions and measured cross-sectional area (CSA) and thickness in September and October 2023. Participants were assessed either at the Musculoskeletal Research Group laboratory (University of Leuven, Bruges) or in the Rehabilitation Sciences laboratory (Ghent University hospital). Test-retest (same investigator, one week in between), global inter-rater (each investigator measures own image set) and MIRR (three investigators measure one image set) was performed following intra-class correlation, standard error of the measurement (SEM) and coefficient of variation.

Results: Test-retest reliability showed intraclass-correlation coefficients of 0.60-0.88 for the FDB and 0.38-0.73 for the TA. SEM ranged from 0.16 to 0.41 cm² (CSA) and from 0.05 to 0.31 cm (thickness) for the intrinsic, while they ranged from 0.19 to 1.13 cm² and from 0.12 to 0.44 cm for the extrinsic muscles. Global inter-rater correlation coefficients varied between 0.4 and 0.8 for the AbH and FDB. Measurement based inter-rater correlation coefficient varied between 0.50 and 0.96 for AbH, FDB, TA and PER muscles. SEM ranged from 0.14 to 0.89 cm² (CSA) and from 0.07 to 0.24 cm (thickness) for the intrinsic, while they ranged from 0.29 to 0.85 cm² (CSA) and from 0.12 to 0.51 cm (thickness) for the extrinsic muscles. Coefficients of variations were between 4% and 34%. For test-retest, they were consistently ≤10% for AbH thickness, FDB CSA, FHB and TA. FDB coefficients of variation were ≤10% across all inter-rater reliabilities.

Conclusions: Most muscles demonstrated moderate to excellent test-retest reliability using a portable ultrasound device, supporting its generalizability. However, the greater variability in global inter-rater reliability suggests substantial variation during image acquisition. The same clinician should perform pre-intervention and follow-up assessments to minimize errors. If different clinicians are involved, caution is needed when comparing measurements.