Tendon integrity can be described using a wide range of biomarkers via specialized analysis software for images recorded by musculoskeletal ultrasonography. This study proposes a minimal biomarker data set to characterize Achilles tendon ultrasound images and to differentiate between symptomatic and asymptomatic Achilles tendon images. Forty-one individuals with unilateral Achilles tendinopathy and 35 healthy controls had their Achilles tendon images recorded bilaterally by ultrasonography in the longitudinal and transverse planes. A set of 22 biomarkers, organized around three dimensions (geometric, composition and texture), were computed in each plane. A symmetry index, reflecting relative side differences, was compared between groups and analyzed through principal component analysis to isolate biomarkers that best explained data variance. A specific minimal data set was identified by linear regression in the longitudinal (mean thickness, echogenicity, variance and homogeneity at 90°) and transverse (mean thickness, echogenicity and mean homogeneity) planes to characterize and differentiate Achilles tendon integrity.
Keywords: Computer-assisted image interpretation; Diagnostic imaging; Image processing; Principal component analysis; Rehabilitation; Tendinopathy; Tendinosis; Ultrasonography.
Copyright © 2020 World Federation for Ultrasound in Medicine & Biology. Published by Elsevier Inc. All rights reserved.