kth.sePublications KTH
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Quantitative Assessment of Muscle Morphology and Texture Features Using 3D Freehand Ultrasound
KTH, School of Engineering Sciences (SCI), Engineering Mechanics.ORCID iD: 0009-0005-1316-3268
2026 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Skeletal muscle plays a central role in mobility, physical function, and overall health. Accurate quantification of skeletal muscle morphology and quality is essential for understanding muscle adaptations associated with impairment and aging, as well as for evaluating rehabilitation outcomes. This thesis investigates both the clinical application and technical advancement of three-dimensional freehand ultrasound (3DfUS) as a comprehensive tool for in vivo assessment of skeletal muscle.

In the first study, 3DfUS combined with quantitative texture analysis was used to examine muscle morphology and tissue quality in individuals with subacute stroke. In this randomized controlled study, tibialis anterior muscle volume (MV) and 3DfUS-derived texture features were evaluated before and after a four-week rehabilitation intervention. Participants were assigned to receive either only conventional rehabilitation (CR) or CR with functional electrical stimulation (FES). The results showed a significant increase in MV on the more-affected side and improved inter-limb asymmetry of MV in the FES group, whereas no significant changes were observed in the CR group. Moreover, alterations in texture features were observed on the less-affected side in the FES group, suggesting changes in muscle tissue heterogeneity. These findings support the potential of 3DfUS and texture analysis to detect muscle remodeling during the early phase of post-stroke rehabilitation.

To improve the clinical feasibility of 3DfUS, a low-cost fiducial marker-based tracking system (FM-3DfUS) was developed and validated in the second study as a practical alternative to conventional motion capture-based setups. FM-3DfUS demonstrated excellent agreement with the reference method for muscle morphological parameters of the rectus femoris muscle, including MV, fascicle length, and pennation angle. The system also showed high inter-operator reliability and inter-session repeatability, reflecting strong reproducibility across users and sessions. These findings support the feasibility of accurate 3D muscle assessment using a more clinically applicable setup, and highlight the potential of FM-3DfUS to facilitate broader implementation of quantitative muscle imaging in both research and clinical settings.

Abstract [sv]

Tillförlitlig kvantifiering av skelettmuskelns morfologi och kvalitet är avgörande för att förstå muskelanpassningar kopplade till funktionsnedsättning och åldrande samt för att utvärdera effekter av rehabilitering. Den här avhandlingen undersöker både klinisk tillämpning och teknisk vidareutveckling av handhållen tredimensionell ultraljudsavbildning  (3DfUS) som ett omfattande verktyg för in vivo-bedömning av skelettmuskulatur.

I den första studien användes 3DfUS i kombination med kvantitativ texturanalys för att undersöka muskelmorfologi och vävnadskvalitet hos personer med subakut stroke. I denna randomiserade kontrollerade studie analyserades tibialis anteriors muskelvolym (MV) och 3DfUS-baserade texturegenskaper före och efter en fyra veckor lång rehabiliteringsintervention hos deltagare som randomiserats till två grupper och som genomgick rehabilitering antingen med eller utan funktionell elektrisk stimulering (FES). Resultaten visade en signifikant ökning av MV på den mer påverkade sidan samt minskad asymmetri i MV mellan benen i FES-gruppen, medan konventionell rehabilitering utan FES inte medförde några signifikanta förändringar i MV. Vidare observerades förändringar i texturegenskaper på den mindre påverkade sidan i FES-gruppen, vilket tyder på förändrad heterogenitet i muskelvävnaden. Dessa resultat visar att 3DfUS-baserad texturanalys har potential att identifiera muskelremodellering under den tidiga fasen av rehabilitering efter stroke.

För att öka den kliniska tillgängligheten av 3DfUS utvecklades och validerades sedan i den andra studien ett lågkostnadsbaserat och portabelt system för handhållen tredimensionell ultraljudsavbildning baserat på referensmarkörer (FM-3DfUS), som ett praktiskt alternativ till konventionella rörelseanalysbaserade system. Det portabla systemet visade mycket god överensstämmelse med referensmetoden för centrala muskelmorfologiska parametrar, inklusive muskelvolym (MV), fascikellängd och pennationsvinkel (ICC = 0,92–0,99) i rectus femoris. Systemet uppvisade dessutom hög interbedömarreliabilitet och god upprepbarhetmellan sessioner, vilket tyder på god reproducerbarhet mellan användare och mättillfällen. Resultaten visar att noggrann tredimensionell muskelbedömning kan uppnås med en förenklad och mer kliniskt genomförbar metod, och understryker potentialen hos FM-3DfUS att främja en bredare användning av kvantitativ muskelavbildning inom både klinisk verksamhet och forskning.

Dessa resultat indikerar att 3DfUS är ett tillförlitligt avbildningsverktyg för såväl forskning som kliniska tillämpningar och möjliggör en detaljerad bedömning av muskelns kvantitet och kvalitet. Genom att integrera metodologisk innovation med klinisk forskning belyser denna avhandling potentialen hos portabla ultraljudsbaserade tekniker att förbättra uppföljningen av muskelrehabilitering och bidra till utvecklingen av riktade rehabiliteringsstrategier. Ytterligare studier behövs för att validera dessa metoder i större populationer och för att koppla bildbaserade mått till funktionella utfall.

Place, publisher, year, edition, pages
Stockholm, Sweden: KTH Royal Institute of Technology, 2026.
Series
TRITA-SCI-FOU ; 2026:13
Keywords [en]
Muscle assessment, quantitative ultrasound, musculoskeletal imaging, stroke rehabilitation, fiducial markers, motion capture, methodological validation
Keywords [sv]
Muskelbedömning, kvantitativt ultraljud, muskuloskeletal avbildning, strokerehabilitering, referensmarkörer, rörelseinspelning, metodologisk validering
National Category
Medical Imaging
Research subject
Engineering Mechanics
Identifiers
URN: urn:nbn:se:kth:diva-381799ISBN: 978-91-8106-634-0 (print)OAI: oai:DiVA.org:kth-381799DiVA, id: diva2:2061188
Presentation
2026-06-08, D3, Lindstedtsvägen 5, Stockholm, 13:00
Opponent
Supervisors
Note

QC260521

Available from: 2026-05-21 Created: 2026-05-20 Last updated: 2026-06-01Bibliographically approved
List of papers
1. Quantitative Assessment of Post-Stroke Muscle Following Functional Electrical Stimulation Using 3D Freehand Ultrasound and Texture Analysis
Open this publication in new window or tab >>Quantitative Assessment of Post-Stroke Muscle Following Functional Electrical Stimulation Using 3D Freehand Ultrasound and Texture Analysis
Show others...
(English)Manuscript (preprint) (Other academic)
Abstract [en]

Stroke often leads to persistent motor impairments accompanied by structural and compositional changes in skeletal muscle, both of which can compromise walking ability and functional mobility. Functional electrical stimulation (FES) has been used to improve foot clearance in the swing phase by delivering electrical pulses to compensate for weak dorsiflexor muscles.  Although FES has emerged as a promising intervention to enhance motor recovery, its effects on morphological adaptation and tissue characteristics of the targeted muscle remain unclear. This study aimed to quantify changes in muscle volume (MV) and ultrasound-derived texture features following a four-week rehabilitation program with or without FES using three-dimensional freehand ultrasound (3DfUS).

Twenty individuals with subacute stroke were initially recruited in a randomized controlled trial, and fourteen with complete ultrasound data were included in the final analysis. Participants were randomly assigned to either a conventional rehabilitation (CR) group undergoing four weeks of gait training (n = 7) or CR combined with FES on the tibialis anterior during gait training (FES group, n = 7). MV was assessed pre- and post-intervention using 3DfUS. Muscle tissue quality was further evaluated using radiomic features from volumes of interest, including echovariation, gray-level co-occurrence matrix (GLCM) entropy, and GLCM contrast. Pre- and post-intervention comparisons were conducted within and between groups.

Compared to the baseline (pre-intervention), the FES group showed an increase in MV of 11.0 ± 11.7 cm³ on the more-affected side (p = 0.04), resulting in improved inter-limb volume symmetry. No significant changes in MV were observed in the CR group, suggesting that FES may support the early preservation of muscle mass in the paretic limb. Texture analysis further revealed increased echovariation, entropy and contrast on the less-affected side in the FES group after intervention (all p < 0.05), indicating greater structural heterogeneity. These changes may reflect early muscle remodeling processes in the subacute phase after stroke, while FES may have helped to preserve muscle quality in the more-affected limb. However, no significant changes were observed in the CR group on either side. 

Overall, these findings indicated that FES may influence both muscle morphology and tissue-level quality during the subacute phase after stroke. The combined assessment of MV and texture features suggests that 3DfUS, together with quantitative texture analysis, enabled a more comprehensive evaluation of early muscle adaptation and may provide additional value for monitoring rehabilitation-related changes in skeletal muscle. However, given the exploratory design of the study and the heterogeneity of this population, future studies with larger cohorts and longer follow-up are required to determine the clinical relevance of FES-related adaptation in muscle morphology and tissue characteristics. 

National Category
Medical Imaging
Identifiers
urn:nbn:se:kth:diva-381768 (URN)
Note

QC 20260521

Available from: 2026-05-20 Created: 2026-05-20 Last updated: 2026-05-21Bibliographically approved
2. Reliability and Validity of Computer Vision-Based 3D Freehand Ultrasound for in vivo Muscle Morphology Quantification
Open this publication in new window or tab >>Reliability and Validity of Computer Vision-Based 3D Freehand Ultrasound for in vivo Muscle Morphology Quantification
Show others...
(English)Manuscript (preprint) (Other academic)
Abstract [en]

Three-dimensional freehand ultrasound (3DfUS) is a valuable tool for quantifying 3D skeletal muscle morphology over a large field of view using a standard 2D probe. However, current 3DfUS systems commonly rely on resource-intensive optical motion capture (mocap) systems for probe tracking, which limits their portability and clinical accessibility. This study therefore presented a low-cost monocular alternative that used a cluster of fiducial markers (FM) to track probe motion. The validity, inter-operator reliability, and inter-session repeatability of key muscle morphological parameters, i.e., muscle volume (MV), fascicle length (FL), and pennation angle (PA), were evaluated in the rectus femoris of ten non-disabled participants. The proposed system, FM-3DfUS, showed excellent agreement with the mocap-based reference for MV and FL (ICC = 0.99), and good-to-excellent agreement for PA (ICC = 0.92), with low measurement error across parameters (SEM MV: 1.30 cm3; FL: 2.17 mm; PA: 0.44°). Inter-operator reliability was excellent for all parameters (ICC = 0.91 – 0.98), while inter-session repeatability was excellent for MV and FL (ICC = 0.99) and good for PA (ICC = 0.85). Trajectory analysis confirmed accurate tracking performance, with translational errors < 2% and rotational errors of approximately 3° relative to the mocap reference. These findings demonstrate that FM-3DfUS can provide valid and reliable measurements of key muscle morphological parameters, supporting its potential as a low-cost alternative to conventional mocap-based 3DfUS systems for muscle assessment in both clinical and research settings.

National Category
Medical Imaging
Identifiers
urn:nbn:se:kth:diva-381797 (URN)
Note

QC 20260521

Available from: 2026-05-20 Created: 2026-05-20 Last updated: 2026-05-21Bibliographically approved

Open Access in DiVA

fulltext(6569 kB)45 downloads
File information
File name FULLTEXT01.pdfFile size 6569 kBChecksum SHA-512
7c028ded300928f12e0ba550e425a0284a7330dde6b3eb8c7814370881fb5fae71ad5bd924a3a58e1b3055603b3d9d58df9e6c37a9d0c47121f0a5634cb66ea3
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Huang, Ruoyu
By organisation
Engineering Mechanics
Medical Imaging

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 565 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf