 Original Contribution RELIABILITY OF PROCESSING 3-D FREEHAND ULTRASOUND DATA TO DEFINE MUSCLE VOLUME AND ECHO-INTENSITY IN PEDIATRIC LOWER LIMB MUSCLES WITH TYPICAL DEVELOPMENT OR WITH SPASTICITY T AGGEDPBRITTA HANSSEN,* ,y,z NATHALIE DE BEUKELAER,* ,y SIMON-HENRI SCHLESS,* ,y,x FRANCESCO CENNI, y,{,║ LYNN BAR-ON,* ,y,# NICKY PEETERS,* ,y,z GUY MOLENAERS,** ,yy ANJA VAN CAMPENHOUT,** ,yy CHRISTINE VAN DEN BROECK, z and KAAT DESLOOVERE* ,y TAGGEDEND * Department of Rehabilitation Sciences, KU Leuven, Leuven, Belgium; y Clinical Motion Analysis Laboratory, University Hospitals Leuven, Pellenberg, Belgium; z Department of Rehabilitation Sciences, Ghent University, Ghent, Belgium; x Motion Analysis and Biofeedback Laboratory, ALYN Paediatric and Rehabilitation Hospital, Jerusalem, Israel; { Department of Mechanical Engineering, KU Leuven, Leuven, Belgium; ║ Laboratory of Kinesiology Willy Taillard, Geneva University Hospitals and University of Geneva, Geneva, Switzerland; # Department of Rehabilitation Medicine, Amsterdam UMC, Amsterdam Movement Sciences, Amsterdam, The Netherlands; ** Department of Development and Regeneration, KU Leuven, Leuven, Belgium; and yy Orthopaedic Section, University Hospitals Leuven, Leuven, Belgium (Received 22 November 2020; revised 23 April 2021; in final from 27 April 2021) Abstract—This investigation assessed the processer reliability of estimating muscle volume and echo-intensity of the rectus femoris, tibialis anterior and semitendinosus. The muscles of 10 typically developing children (8.15 [1.40] y) and 15 children with spastic cerebral palsy (7.67 [3.80] y; Gross Motor Function Classification System I = 5, II = 5, III = 5) were scanned with 3-D freehand ultrasonography. For the intra-processer analysis, the intra-class correlations coefficients (ICCs) for muscle volume ranged from 0.9430.997, with relative standard errors of measurement (SEM%) ranging from 1.24%8.97%. For the inter-processer analysis, these values were 0.853 to 0.988 and 3.47% to 14.02%, respectively. Echo-intensity had ICCs >0.947 and relative SEMs <4% for both analyses. Muscle volume and echo-intensity can be reliably extracted for the rectus femoris, semitendi- nosus and tibialis anterior in typically developing children and children with cerebral palsy. The need for a single processer to analyze all data is dependent on the size of the expected changes or differences. (E-mail: Britta. hanssen@kuleuven.be) © 2021 World Federation for Ultrasound in Medicine & Biology. All rights reserved. Key Words: 3-D freehand ultrasonography, Reliability, Muscle volume, Echo-intensity, Cerebral palsy, Children. INTRODUCTION Size and composition of skeletal muscles are important characteristics related to the force-producing capacity of the muscle (Fukunaga et al. 2001). Muscle tissue is very adaptive, and in vivo imaging can be used to evaluate adaptations to positive and negative stimuli such as treat- ment, training, aging and disuse (Fry et al. 2004; McNee et al. 2009; Barber et al. 2013). Therefore, evalu- ation of the morphological characteristics is relevant for both research and clinical practice. Although magnetic resonance imaging (MRI) has been considered the gold standard for muscle volume (MV) evaluations, 3-D freehand ultrasonography (3-DfUS) has become an established alternative method (Cenni et al. 2016; Mozaffari and Lee 2017). Moreover, indirect information on intrinsic muscle composition can be derived from the echo intensity (EI) of ultrasound images, which has been related to the content of contractile and non-contractile tissue in the muscle (Pitcher et al. 2015; Young et al. 2015). Three-dimensional fUS is an imaging technique combining conventional 2-D B-mode ultrasonography with a motion tracking system. The position and orienta- tion of the US probe at every image can be combined to create 3-D data sets of anatomical volumes. By manually drawing transverse plane segmentations along the mus- cle border in the 2-D images and linearly interpolating these cross-sectional areas, a 3-D reconstruction of a muscle can be created from which the MV and average EI can be calculated (Cenni et al. 2016). It provides Address correspondence to: Britta Hanssen, Weligerveld 1, 3212 Pellenberg, Belgium. E-mail: Britta.hanssen@kuleuven.be 1 ARTICLE IN PRESS Ultrasound in Med. & Biol., Vol. 00, No. 00, pp. 111, 2021 Copyright © 2021 World Federation for Ultrasound in Medicine & Biology. All rights reserved. Printed in the USA. All rights reserved. 0301-5629/$ - see front matter https://doi.org/10.1016/j.ultrasmedbio.2021.04.028