In-field assessment of athletic performance and fatigue onset detection during ice skating using wearable sensors

• Author / Creator

  Khandan, Aminreza

• Ice skating requires high aerobic and anaerobic fitness levels, well-coordinated body motion, and efficient neuromuscular systems functioning, requiring continuous performance assessment. Inertial measurement units (IMUs), widely used in tracking human motion for medical diagnostics and clinical evaluations, offer significant potential for enhancing athletic performance. Therefore, we developed and validated a wearable IMU technology for on-ice performance assessment, enabling us to assess skating biomechanics across various ice skating modalities in real-world settings.

  First, we validated the novel algorithms to estimate skating temporal and spatial parameters by proposing an optimized configuration of wearable IMUs. Ten participants were recruited to skate on a 14-m synthetic ice surface built in a motion-capture lab. Stride time, contact time, stride length, and stride velocity were obtained with a 2-6\% relative error compared to the in-lab motion capture reference system. We demonstrated that our wearable IMU technology on skates and pelvis could accurately and precisely estimate skating temporal and spatial parameters with similar relative errors compared to those obtained in IMU-based gait analysis.

  Second, we explored the effectiveness of the on-ice distinctive features measured using these wearable sensors in differentiating low- and high-calibre skaters. Six high-calibre and six low-calibre skaters were recruited to skate forward on a synthetic ice surface. Five IMUs were placed on their dominant leg and pelvis. The 3D lower-limb joint angles obtained by IMUs showed a maximum root mean square error of 5 degrees against those obtained by a motion capture system. Our findings indicated that synthetic ice experiments impact skating 3D joint angles, blurring the differences between low- and high-calibre skaters typically seen in on-ice skating.

  Third, we showed the potential of our wearable technology to track skaters' performance, predict perceived fatigue, and detect the onset of severe fatigue. In multistage aerobic experiments, nineteen high- and low-calibre skaters clustered by our proposed algorithm skated at a self-selected speed around an ice rink. Our developed algorithms measured 22 kinematic metrics using IMUs mounted on the dominant lower limb. The variations of inter-segment angle correlation, joint angle fluctuations, and trunk angle with perceived fatigue during aerobic ice skating were considerable. Finally, using the proposed kinematic metrics, we employed a gradient-boosting machine learning model to predict severe fatigue onset with high performance.

  Fourth, we assessed skating performance using an expanded range of performance metrics obtained from our wearable technology in forward ice sprint tests. Nineteen ice skaters were recruited to sprint on ice with maximal speed while six IMUs recorded their movements. We found that stride velocity and stride length differed between low- and high-calibre skaters, and stride velocity differed between figure and hockey skaters. Also, figure skaters skated less complex and more coordinated than hockey players during the tests. Finally, we showed that the metrics we introduced can guide further research on exploring additional suitable off-ice tests, enabling the prediction of on-ice performance through off-ice measurements.

  The outcome of this thesis research is a user-friendly wearable sensor system to provide an accurate outlook for ice skating coaches to improve their tutoring methods and youth/adult athletes' learning outcomes. This wearable technology demonstrated a significant potential to deepen our understanding of skating biomechanics and offer valuable insights for enhancing skating performance in multistage aerobic and sprint tests. Future studies could broaden the scope of this technology to include different skating styles and specific applications in hockey matches. Furthermore, IMU-based evaluations have indicated a potential for early detection of fatigue, aiming to reduce fatigue-induced injury risks in skaters of different calibres.

• Subjects / Keywords

  • wearable inertial measurement technology
  • ice skating biomechanics
  • skating performance assessment
  • performance fatigue detection
  • ice skating sprint test
  • synthetic ice skating
  • kinematic metrics
  • temporal and spatial parameters
  • aerobic intermittent ice skating
  • clustering algorithm
  • three-dimensional kinematics
  • perceived fatigue
  • athletic performance assessment
  • machine learning models in sports

• Graduation date

  Fall 2024

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-rnas-ba21

• License

  This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.