1 / 9

Predicting Post-Operative Patient Gait

Predicting Post-Operative Patient Gait . Jongmin Kim Movement Research Lab. Seoul National University. Pose predictor. Learn a pose predictor from training data set . - : pre-operative patient’ pose (input) - : post-operative patient’ pose (output)

ariane
Download Presentation

Predicting Post-Operative Patient Gait

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Predicting Post-Operative Patient Gait Jongmin Kim Movement Research Lab. Seoul National University

  2. Pose predictor • Learn a pose predictor from training data set . - : pre-operative patient’ pose (input) - : post-operative patient’ pose (output) • Given new input data, we generate a new pose using the learned predictor. Regression process Predictor New input data, x Output pose Motion database

  3. L1 regularization • Regularization based on the L1 drives maximizes sparseness. • A new predicting post-operative gait can be estimated as matrix-vector multiplication. - e.g. L1 sparsity term

  4. The problems • Correspondence • Can not explain the nonlinear relationship between training input and output. Pre-operative patient’s motion Post-operative patient’s motion

  5. Mapping model • Motion to motion - handling correspondence between remote poses in temporal space • Find the method that fully explains the nonlinear relationship between training input and output. - Nonlinear statistical regression methods

  6. Related work • Predicting outcomes of rectus femoris transfer surgery [Reinbolt et al. 2009] • Long term outcome of single event multilevel surgery in spastic diplegia with flexed knee gait [Sung et al. 2012]

  7. Design features • Kinematic features [Sung et al. 2012] - mean pelvic tilt - minimum hip flexion - … • Middle swing left - knee flexion - left hip flexion - ankle plantarflexion right leg left leg

  8. Design features • Left toe-off - knee flexion (velocity) - ankle plantarflexion (velocity) - right hip flexion - pelvis forward velocity • Terminal swing left - knee flexion - ankle plantarflexion - left hip internal rotation left leg right leg right leg left leg

  9. Future work • Design training input & output with respect to the clinical context. (very important !) - GCD data • Motion to motion • Nonlinear statistical regression models - SVR, Gaussian process, …

More Related