Back to available technologies

Systems and methods for the automated assessment of athletic performance and movement quality

Value Proposition

Risk of lower body injury in athletes, such as an ACL tear, can be reliably predicted through clinical assessment of jump-landing biomechanics. However, such assessments must be administered by trained professionals, and the results of these tests are often subjective. These restrictions hinder widespread adoption and routine use of jump-landing tests, thus limiting the overall potential for injury prevention. The present technology provides a computational method for the automated assessment and risk classification of athletes performing a jump-landing test. Risk of lower body injury in athletes, such as an ACL tear, can be reliably predicted through clinical assessment of jump-landing biomechanics. However, such assessments must be administered by trained professionals, and the results of these tests are often subjective. These restrictions hinder widespread adoption and routine use of jump-landing tests, thus limiting the overall potential for injury prevention. The present technology provides a computational method for the automated assessment and risk classification of athletes performing a jump-landing test.

Technology

Through the use of commonly-available wearable accelerometers and / or force plates, this technology uses a novel machine-learning approach to accurately predict whether an athlete is at risk for injury, enabling pre-emptive treatment and preventing injury. As an automated system, this technology can reduce costs and expand availability of testing, enabling previously unfeasible approaches such as longitudinal monitoring of athletes following corrective therapy or training. Using only 40 participants to train the model, this approach showed accuracy of 80% (accelerometer) or 87.5% (force plate) relative to standard professional assessment. It is expected that this accuracy can be substantially increased through incorporation of additional parameters and further training of the algorithm on additional participants.

Other Applications

This technology could potentially be adapted to accept input from existing wearable fitness monitors to provide real-time feedback to athletes looking to improve biomechanics during training.

Advantages

  • Automated approach is much more scalable than human assessment by a trained professional.
  • Enables longitudinal assessment, as well as self-assessment requiring minimal equipment.

Duke File (IDF) Number

T-005307

Inventor(s)

  • Mazzoleni, Michael
  • Frank, Barnett
  • Mann, Brian
  • Padua, Darin

For more information please contact

College

Pratt School of Engineering

Interested in this Technology?

Submit your interest below.