Hiring in life sciences? Share your open positions with our professional community. Read more Close

Advertisement

BoneNet: a framework for predicting MSK-FE-estimated femoral neck strain patterns from exercise IMU data.

Created on 16 Nov 2026

Authors

Zainab Altai, David Hollinger, Andrew T M Phillips, Jason Moran, Xiaojun Zhai, Qichang Mei, Bernard X W Liew

Published in

Journal of biomechanics. Volume 210. Pages 113677. Nov 11, 2026. Epub Nov 11, 2026.

Abstract

Femoral neck fractures pose serious risks, particularly in osteoporosis, highlighting the need for effective bone-strengthening exercises and practical, non-invasive assessment of femoral loading. This study quantifies exercise-induced strains and presents BoneNet, an explainable convolutional neural network that predicts femoral neck strains from inertial measurement unit (IMU) data, offering a potential approach for estimating musculoskeletal-finite element (MSK-FE)-derived femoral neck strain patterns. Data were collected from 40 healthy participants (40.3 ± 13.1 years), with 37 included in the FE and BoneNet analyses. MSK-FE models estimated femoral neck strains during walking, running, jumping and hopping at varying intensities. BoneNet used IMU accelerometer and gyroscope signals from seven- and three-sensor configurations with participant-grouped cross-validation. Fast running yielded 3731 ± 1472 µε (tensile) and - 9541 ± 3610 µε (compressive), with walking strain magnitudes 47% lower for both tensile and compressive strains. Unilateral hopping with longer ground contact yielded 3003 ± 1189 µε and - 7755 ± 3040 µε, with walking strain magnitudes 34% and 35% lower, respectively. Walking generally produced higher strains than all other exercises. Using seven IMUs, BoneNet achieved correlations of r = 0.75-0.94 for five exercises, while squat jumps showed weak correlations (r = 0.27-0.36) and high relative RMSE (54.9-59.4%). Three IMUs showed comparable correlations, but squat-jump errors remained high (relative RMSE 56.0-61.2%). Running at all intensities and unilateral hopping with longer ground contact may have implications for bone adaptation, while BoneNet shows potential for estimating strain patterns rather than validated absolute values.

PMID:
43035064
Bibliographic data and abstract were imported from PubMed on 16 Nov 2026.

Read full publication at:
Please sign in to see all details.

Advertisement

Stats

  • Community rating n/a 0 votes
  • Reviewers' rating n/a 0 votes
  • Your rating

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

  • Recommendations n/a n/a positive of 0 vote(s)
  • Views 8
  • Comments 0

Recommended by

  • No recommendations yet.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.

Advertisement