Andrea Sabo1,2, Carolina Gorodetsky3,4, Alfonso Fasano1,4,5,6,7, Andrea Iaboni1,8,9, Babak Taati1,2,10. 1. KITE Research Institute, Toronto Rehabilitation Institute-University Health Network (UHN) Toronto ON M5G 1L7 Canada. 2. Institute of Biomedical Engineering, University of Toronto Toronto ON M5S 1A1 Canada. 3. Division of NeurologyThe Hospital for Sick Children Toronto ON M5G 1X8 Canada. 4. Edmond J. Safra Program in Parkinson's Disease, Morton and Gloria Shulman Movement Disorders Clinic, Toronto Western HospitalUniversity Health Network (UHN) Toronto ON M5G 1L7 Canada. 5. Division of NeurologyUniversity of Toronto Toronto ON M5S 1A1 Canada. 6. Krembil Brain Institute Toronto ON M5T 1M8 Canada. 7. Center for Advancing Neurotechnological Innovation to Application (CRANIA) Toronto ON M5G 2C4 Canada. 8. Department of PsychiatryUniversity of Toronto Toronto ON M5S 1A1 Canada. 9. Centre for Mental HealthUniversity Health Network (UHN) Toronto ON M5G 1L7 Canada. 10. Department of Computer ScienceUniversity of Toronto Toronto ON M5S 1A1 Canada.
Abstract
BACKGROUND: Parkinson's disease (PD) presents with motor symptoms such as bradykinesia, rigidity, and tremor that can affect gait. To monitor changes associated with disease progression or medication use, quantitative gait assessment is often performed during clinical visits. Conversely, vision-based solutions have been proposed for monitoring gait quality in non-clinical settings. METHODS: We use three 2D human pose-estimation libraries (AlphaPose, Detectron, OpenPose) and one 3D library (ROMP) to calculate gait features from color video, and correlate them with those extracted by a Zeno instrumented walkway in older adults with PD. We calculate video-based gait features using a manual and automated heel-strike detection algorithm, and compare the correlations when the participants walk towards and away from the camera separately. RESULTS: Based on analysis of 67 bidirectional walking bouts from 25 adults with PD, moderate to strong positive correlations were identified between the number of steps, cadence, as well as the mean and coefficient of variation of step width calculated from Zeno and video using 2D pose-estimation libraries. We noted that our automated heel-strike annotation method struggled to identify short steps. CONCLUSION: Gait features calculated from 2D joint trajectories are more strongly correlated with the Zeno than analogous gait features calculated from ROMP. Based on our analysis, videos processed with 2D pose-estimation libraries can be used for longitudinal gait monitoring in individuals with PD. Future work will seek to improve the prediction of gait features using a comprehensive machine learning model to predict gait features directly from color video without relying on intermediate extraction of joint trajectories.
BACKGROUND: Parkinson's disease (PD) presents with motor symptoms such as bradykinesia, rigidity, and tremor that can affect gait. To monitor changes associated with disease progression or medication use, quantitative gait assessment is often performed during clinical visits. Conversely, vision-based solutions have been proposed for monitoring gait quality in non-clinical settings. METHODS: We use three 2D human pose-estimation libraries (AlphaPose, Detectron, OpenPose) and one 3D library (ROMP) to calculate gait features from color video, and correlate them with those extracted by a Zeno instrumented walkway in older adults with PD. We calculate video-based gait features using a manual and automated heel-strike detection algorithm, and compare the correlations when the participants walk towards and away from the camera separately. RESULTS: Based on analysis of 67 bidirectional walking bouts from 25 adults with PD, moderate to strong positive correlations were identified between the number of steps, cadence, as well as the mean and coefficient of variation of step width calculated from Zeno and video using 2D pose-estimation libraries. We noted that our automated heel-strike annotation method struggled to identify short steps. CONCLUSION: Gait features calculated from 2D joint trajectories are more strongly correlated with the Zeno than analogous gait features calculated from ROMP. Based on our analysis, videos processed with 2D pose-estimation libraries can be used for longitudinal gait monitoring in individuals with PD. Future work will seek to improve the prediction of gait features using a comprehensive machine learning model to predict gait features directly from color video without relying on intermediate extraction of joint trajectories.
Authors: Kimberley-Dale Ng; Sina Mehdizadeh; Andrea Iaboni; Avril Mansfield; Alastair Flint; Babak Taati Journal: IEEE J Transl Eng Health Med Date: 2020-05-28 Impact factor: 3.316
Authors: Christopher G Goetz; Barbara C Tilley; Stephanie R Shaftman; Glenn T Stebbins; Stanley Fahn; Pablo Martinez-Martin; Werner Poewe; Cristina Sampaio; Matthew B Stern; Richard Dodel; Bruno Dubois; Robert Holloway; Joseph Jankovic; Jaime Kulisevsky; Anthony E Lang; Andrew Lees; Sue Leurgans; Peter A LeWitt; David Nyenhuis; C Warren Olanow; Olivier Rascol; Anette Schrag; Jeanne A Teresi; Jacobus J van Hilten; Nancy LaPelle Journal: Mov Disord Date: 2008-11-15 Impact factor: 10.338
Authors: Ana Lígia Silva de Lima; Tim Hahn; Luc J W Evers; Nienke M de Vries; Eli Cohen; Michal Afek; Lauren Bataille; Margaret Daeschler; Kasper Claes; Babak Boroojerdi; Dolors Terricabras; Max A Little; Heribert Baldus; Bastiaan R Bloem; Marjan J Faber Journal: PLoS One Date: 2017-12-20 Impact factor: 3.240
Authors: Johannes C M Schlachetzki; Jens Barth; Franz Marxreiter; Julia Gossler; Zacharias Kohl; Samuel Reinfelder; Heiko Gassner; Kamiar Aminian; Bjoern M Eskofier; Jürgen Winkler; Jochen Klucken Journal: PLoS One Date: 2017-10-11 Impact factor: 3.240