High variability in strain estimation errors when using a commercial ultrasound speckle tracking algorithm on tendon tissue
2016 (English)In: Acta Radiologica, ISSN 0284-1851, E-ISSN 1600-0455, Vol. 57, no 10, 1223-1229 p.Article in journal (Refereed) Published
Background: Ultrasound speckle tracking offers a non-invasive way of studying strain in the free Achilles tendon where no anatomical landmarks are available for tracking. This provides new possibilities for studying injury mechanisms during sport activity and the effects of shoes, orthotic devices, and rehabilitation protocols on tendon biomechanics. Purpose: To investigate the feasibility of using a commercial ultrasound speckle tracking algorithm for assessing strain in tendon tissue. Material and Methods: A polyvinyl alcohol (PVA) phantom, three porcine tendons, and a human Achilles tendon were mounted in a materials testing machine and loaded to 4% peak strain. Ultrasound long-axis cine-loops of the samples were recorded. Speckle tracking analysis of axial strain was performed using a commercial speckle tracking software. Estimated strain was then compared to reference strain known from the materials testing machine. Two frame rates and two region of interest (ROI) sizes were evaluated. Results: Best agreement between estimated strain and reference strain was found in the PVA phantom (absolute error in peak strain: 0.21 +/- 0.08%). The absolute error in peak strain varied between 0.72 +/- 0.65% and 10.64 +/- 3.40% in the different tendon samples. Strain determined with a frame rate of 39.4Hz had lower errors than 78.6Hz as was the case with a 22mm compared to an 11mm ROI. Conclusion: Errors in peak strain estimation showed high variability between tendon samples and were large in relation to strain levels previously described in the Achilles tendon.
Place, publisher, year, edition, pages
Sage Publications, 2016. Vol. 57, no 10, 1223-1229 p.
Speckle tracking, strain, Achilles tendon, ultrasound
Medical Equipment Engineering Medical Laboratory and Measurements Technologies
IdentifiersURN: urn:nbn:se:kth:diva-193792DOI: 10.1177/0284185115626471ISI: 000382967500013PubMedID: 26787677ScopusID: 2-s2.0-84987786849OAI: oai:DiVA.org:kth-193792DiVA: diva2:1039535
FunderStockholm County CouncilSwedish National Centre for Research in Sports
QC 201610242016-10-242016-10-112016-10-24Bibliographically approved