IP Library Granted Patent US 12,635,907
Granted Patent B2
US 12,635,907 · App. 17/260,970 · Granted May 26, 2026

System and method for automatic evaluation of gait using single or multi-camera recordings

Inventors: Scott Delp (Stanford, CA); Lukasz Kidzinski (Menlo Park, CA); Bryan Yang (San Jose, CA); Michael Schwartz (Minneapolis, MN); Jennifer Lee Hicks (Fremont, CA)
Assignees: The Board of Trustees of the Leland Stanford Junior University; Gillette Children's Specialty Healthcare
A61B5/112A61B5/1128A61B5/4082A61B5/7267G16H50/20G16H50/30G16H50/50A61B2576/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,635,907
App. No.
17/260,970
Granted
May 26, 2026
Kind
B2
Abstract

Systems and methods in accordance with many embodiments of the invention include a motion evaluation system that trains a model to evaluate motion (such as, but not limited to, gait) through images (or video) captured by a single image capture device. In certain embodiments, motion evaluation includes predicting clinically relevant variables from videos of patients walking from keypoint trajectories extracted from the captured images.

Claims (44)

1 . A method for evaluating motion from a video, the method comprising:

identifying a set of one or more keypoint trajectories from each of a plurality of frames of a video;

providing the set of one or more keypoint trajectories as inputs to an artificial neural network,

wherein the artificial neural network is trained using multivariate time series of coordinates of a set of one or more keypoint trajectories extracted from at least one training video, wherein each of the at least one training video is annotated with a true motion evaluation score,

wherein the artificial neural network is trained to predict motion evaluation scores based on the multivariate time series of coordinates and the true motion evaluation score of the corresponding training video where the multivariate time series of coordinates are extracted from,

wherein the artificial neural network comprises a convolutional neural network having one or more convolutional layers, wherein each of the one or more convolutional layers is configured to:

receive as input, a multivariate time series of coordinates structured as a T×D matrix, where T represents a number of points in a time dimension of the multivariate time series of coordinates, and D represents a depth dimension of the multivariate time series of coordinates; and

apply one or more filters of a filter length F to the multivariate time series of coordinates, wherein each filter connects to a local region in the time dimension comprising F consecutive points in the time dimension, and wherein each filter extends through an entirety of the depth dimension such that each filter connects to all D features at each of the F consecutive points in the time dimension;

computing the motion evaluation score based on the set of one or more keypoint trajectories using the artificial neural network; and

providing an output based on the motion evaluation score.

2 . The method of claim 1 , wherein identifying the set of keypoint trajectories comprises:

identifying a set of keypoints within each of a plurality of frames of the video;

computing each keypoint trajectory of the set of keypoint trajectories based on positions of the keypoint in each frame of the plurality of frames.

3 . The method of claim 2 , wherein the set of keypoints comprises two-dimensional (2D) positions of joints and body keypoints of an individual captured in the video.

4 . The method of claim 3 , wherein identifying the set of keypoints comprises using an OpenPose process to identify the set of keypoints.

5 . The method of claim 1 , wherein identifying the set of keypoint trajectories further comprises computing additional features from the identified set of keypoints.

6 . The method of claim 1 , wherein the motion evaluation score is one of gait deviation index (GDI), walking speed, cadence, symmetry, gait variability, and stride length.

7 . The method of claim 1 , wherein providing the output comprises providing a treatment regimen for a patient, based on the motion evaluation score.

8 . The method of claim 1 , wherein providing the output comprises providing a diagnosis for a disease.

9 . The method of claim 8 , wherein the disease is one of Parkinson's disease, osteoarthritis, stroke, cerebral palsy, multiple sclerosis, and muscular dystrophy.

10 . The method of claim 1 , wherein providing the output comprises providing the output for an individual's progression based on a plurality of predicted motion evaluation scores over a period of time.

11 . The method of claim 1 , wherein providing the output comprises providing real-time feedback to a user to adjust the user's motions.

12 . The method of claim 1 , wherein, the method further comprising:

performing a physics-based simulation based on the set of keypoint trajectories; and

training a model based on the physics-based simulation, wherein predicting the motion evaluation score comprises using the trained model to predict the motion evaluation score, wherein the motion evaluation score comprises at least one of muscle activation, muscle fiber length, and joint loads.

13 . A non-transitory machine readable medium containing processor instructions for evaluating motion from a video, where execution of the instructions by a processor causes the processor to perform a process that comprises:

identifying a set of one or more keypoint trajectories from a plurality of frames of a video;

providing the set of one or more keypoint trajectories as inputs to an artificial neural network,

wherein the artificial neural network is trained using multivariate time series of coordinates of a set of one or more keypoint trajectories extracted from at least one training video, wherein each of the at least one training video is annotated with a true motion evaluation score,

wherein the artificial neural network is trained to predict motion evaluation scores based on the multivariate time series of coordinates and the true motion evaluation score of the corresponding training video where the multivariate time series of coordinates are extracted from,

wherein the artificial neural network comprises a convolutional neural network having one or more convolutional layers, wherein each of the one or more convolutional layers is configured to:

receive as input, the multivariate time series of coordinates structured as a T×D matrix, where T represents a number of points in a time dimension of the multivariate time series of coordinates, and D represents a depth dimension of the multivariate time series of coordinates; and

apply one or more filters of a filter length F to the multivariate time series of coordinates, wherein each filter connects to a local region in the time dimension comprising F consecutive points in the time dimension, and wherein each filter extends through an entirety of the depth dimension such that each filter connects to all D features at each of the F consecutive points in the time dimension;

computing the motion evaluation score based on the set of one or more keypoint trajectories using the artificial neural network; and

providing an output based on the motion evaluation score.

14 . The non-transitory machine readable medium of claim 13 , wherein identifying the set of keypoint trajectories comprises:

identifying a set of keypoints within each of a plurality of frames of the video;

computing each keypoint trajectory of the set of keypoint trajectories based on positions of the keypoint in each frame of the plurality of frames.

15 . The non-transitory machine readable medium of claim 13 , wherein identifying the set of keypoint trajectories further comprises computing additional features from the identified set of keypoints.

16 . The non-transitory machine readable medium of claim 13 , wherein the motion evaluation score is one of gait deviation index (GDI), walking speed, cadence, symmetry, gait variability, and stride length.

17 . The non-transitory machine readable medium of claim 13 , wherein providing the output comprises providing the output for an individual's progression based on a plurality of predicted motion evaluation scores over a period of time.

18 . The non-transitory machine readable medium of claim 13 , wherein, the non-transitory machine readable medium, wherein the process further comprises:

performing a physics-based simulation based on the set of keypoint trajectories; and

training a model based on the physics-based simulation, wherein predicting the motion evaluation score comprises using the trained model to predict the motion evaluation score, wherein the motion evaluation score comprises at least one of muscle activation, muscle fiber length, and joint loads.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 9, 2021
From: DELP, SCOTT; KIDZINSKI, LUKASZ; YANG, BRYAN; HICKS, JENNIFER LEE
To: THE BOARD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIVERSITY
Reel/Frame 057125/0968 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 9, 2021
From: SCHWARTZ, MICHAEL
To: GILLETTE CHILDREN'S SPECIALTY HEALTHCARE
Reel/Frame 057126/0057 →
Continuity (2)
Provisional Application 62698725 · Jul 16, 2018
Related Publication 20210315486A1 · Oct 14, 2021
References Cited (47)
US 7257237B1 · Luck et al. · 2007 [cited by applicant]
US 9445769B2 · Ghassemzadeh et al. · 2016 [cited by applicant]
US 9604142B2 · Bentley et al. · 2017 [cited by applicant]
US 9700241B2 · Eastman et al. · 2017 [cited by applicant]
US 9811720B2 · Zhong · 2017 [cited by applicant]
US 20070229522A1 · Wang et al. · 2007 [cited by applicant]
US 20120253201A1 · Reinhold · 2012 [cited by applicant]
US 20150106024A1 · Lightcap · 2015 [cited by examiner]
US 20170000386A1 · Salamatian et al. · 2017 [cited by applicant]
US 20180020954A1 · Lillie et al. · 2018 [cited by applicant]
US 20190110736A1 · Broderick · 2019 [cited by examiner]
US 20190110754A1 · Rao · 2019 [cited by examiner]
US 20210307621A1 · Svenson · 2021 [cited by examiner]
CA 2466210C · 2014 [cited by applicant]
CN 111144217A · 2020 [cited by applicant]
CN 110458046B · 2020 [cited by applicant]
TW 202016691A · 2020 [cited by applicant]
WO 2020018469A1 · 2020 [cited by applicant]
International Search Report and Written Opinion for International Application No. PCT/US2019/041896, Search completed Sep. 17, 2019, Mailed Oct. 2, 2019, 13 Pgs. [cited by applicant]
“Gait Analysis”, UnitedHeatlhcare Commercial Medical Policy, May 1, 2017, 9 pgs. [cited by applicant]
Balazia et al., “Gait Recognition from Motion Capture Data”, ACM Transactions on Multimedia Computing, Communications, and Applications, vol. 1, No. 1, Oct. 2017, 18 pgs., https://doi.org/10.1145/nnnnnnn.nnnnnnn. [cited by applicant]
Balazia et al., “Learning Robust Features for Gait Recognition by Maximum Margin Criterion”, Proceedings of the 23rd International Conference on Pattern Recognition, Cancun, Mexico, Dec. 4-8, 2016, 6 pgs., arXiv:1609.04… [cited by applicant]
Benndorf et al., “Automated Annotation of Sensor data for Activity Recognition using Deep Learning”, INFORMATIK 2017, Gesellschaft fur Informatik, Bonn 2017, pp. 2211-2219, doi:10.18420/in2017_220. [cited by applicant]
Cao et al., “Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields”, arXiv:1611.08050v2 [cs.CV], Apr. 14, 2017, 9 pgs. [cited by applicant]
Castelli et al., “A 2D Markerless Gait Analysis Methodology: Validation on Healthy Subjects”, Hindawi Publishing Corporation, Computational and Mathematical Methods in Medicine, vol. 2015, No. 186780, 11 pgs., http://dx… [cited by applicant]
Castro et al., “Evaluation of CNN Architectures for Gait Recognition Based on Optical Flow Maps”, Proceedings of the International Conference of the Biometrics Special Interest Group (BIOSIG), Darmstadt, Germany, Sep. 2… [cited by applicant]
Chau, “A review of analytical techniques for gait data. Part 2: neural network and wavelet methods”, Gait & Posture, vol. 13, Apr. 2001, pp. 102-120. [cited by applicant]
Crabbe et al., “Skeleton-Free Body Pose Estimation from Depth Images for Movement Analysis”, Proceedings of the IEEE International Conference on Computer Vision Workshop (ICCVW), Santiago, Chile, Dec. 7-13, 2015, pp. 70… [cited by applicant]
Dehzangi et al., “IMU-Based Gait Recognition Using Convolutional Neural Networks and Multi-Sensor Fusion”, Sensors, vol. 17, No. 12, Nov. 27, 2017, 22 pgs., doi:10.3390/d17122735. [cited by applicant]
Devanne et al., “Learning Shape Variations of Motion Trajectories for Gait Analysis”, Proceedings of the 23rd International Conference on Pattern Recognition (ICPR), Cancun, Mexico, Dec. 4-8, 2016, pp. 895-900, doi:10.1… [cited by applicant]
Feng et al., “Learning Effective Gait Features Using LSTM”, Proceedings of the 23rd International Conference on Pattern Recognition (ICPR), Cancun, Mexico, Dec. 4-8, 2016, pp. 320-325. [cited by applicant]
Goffredo et al., “2D Markerless Gait Analysis”, Proceedings of the 4th European Conference of the International Federation for Medical and Biological Engineering, vol. 22, 2008, pp. 67-71. [cited by applicant]
Hamill et al., “Read a Chapter on Angular Kinematics”, Angular Kinematics, 20 pgs. [cited by applicant]
Hinton et al., “Lecture 6”, Neural Networks for Machine Learning, Lecture 6A-6E, 31 pgs. [cited by applicant]
Kaczmarczyk et al., “Artificial Neural Networks (ANN) Applied for Gait Classification and Physiotherapy Monitoring in Post Stroke Patients”, InTech, Chapter 16 of Artificial Neural Networks—Methodological Advances and B… [cited by applicant]
Ke et al., “A New Representation of Skeleton Sequences for 3D Action Recognition”, arXiv:1703.03492v3 [cs.CV], Jun. 5, 2017, 10 pgs. [cited by applicant]
Li et al., “Skeleton-based Action Recognition Using LSTM and CNN”, arXiv:1707.02356v1 [cs.CV], Jul. 6, 2017, 6 pgs. [cited by applicant]
Liu et al., “Learning Efficient Spatial-Temporal Gait Features with Deep Learning for Human Identification”, Neuroinformatics, Oct. 2018, vol. 16, No. 3-4, Online Publication: Feb. 6, 2018, pp. 457-471, https://doi.org/… [cited by applicant]
Mallick, “Deep Learning based Human Pose Estimation using OpenCV (C++ / Python)”, IEEE Transactions on Computing, May 29, 2018, vol. 58, No. 10, pp. 1-12. [cited by applicant]
Maudsley-Barton et al., “A Comparative Study of the Clinical use of Motion Analysis from Kinect Skeleton Data”, arXiv:1707.08813v2 [cs.CV], Jul. 31, 2017, 6 pgs. [cited by applicant]
Paiement et al., “Online quality assessment of human movement from skeleton data”, In Proceedings British Machine Vision Conference, Sep. 2014, 12 pgs. [cited by applicant]
Sokolova et al., “Gait Recognition Based on Convolutional Neural Networks”, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XLII-2/W4, 2017, pp. 207-212, doi:10.51… [cited by applicant]
Sokolova et al., “Pose-based Deep Gait Recognition”, arXiv:1710.06512v3 [cs.CV], Feb. 8, 2018, 10 pgs. [cited by applicant]
Speciali et al., “Use of the Gait Deviation Index and spatiotemporal variables for the assessment of dual task interference paradigm”, Journal of Bodywork and Movement Therapies, vol. 17, No. 1, Jan. 2013, pp. 19-27. [cited by applicant]
Tanawongsuwan et al., “Gait Recognition from Time-normalized Joint-angle Trajectories in the Walking Plane”, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Kauai, Hawaii,… [cited by applicant]
Zadeh et al., “Abnormal Walking Gait Analysis Using neural network”, Australian Journal of Basic and Applied Sciences, vol. 5, No. 9, Sep. 2011, pp. 1381-1390. [cited by applicant]
International Preliminary Report on Patentability for International Application PCT/US2019/041896, Issued on Jan. 19, 2021, Mailed on Jan. 28, 2021, 7 pages. [cited by applicant]