Presenter Type
UNO Graduate Student (Doctoral)
Major/Field of Study
Biomechanics
Other
Biomechanics
Author ORCID Identifier
0000-0001-9915-6950
Advisor Information
Assistant Professor
Location
CEC RM #128
Presentation Type
Oral Presentation
Start Date
22-3-2024 2:30 PM
End Date
22-3-2024 3:45 PM
Abstract
Most humans have a fingerprint that is unique and persists throughout life. The same may be true for gait. Each person is unique physiologically and biomechanically, and has experienced a unique progression of life including injuries, habits, hobbies, or lifestyles that all contribute to navigating the world. Those individualized experiences, whether intended or not, are potentially on display in gait. We hypothesized that there are gait characteristics intrinsic and unique to everyone, so that everyone has a unique “gaitprint”, similar to humans possessing unique fingerprints. To test our hypothesis, we recruited thirty healthy young adults between the ages 19-35 as part of the NONAN GaitPrint project. In this study we put simple gait features through several distance and machine learning algorithms to identify people based on how they walk. Each participant completed 18, 4-minute walking trials at a self-selected pace on an indoor track. The motion of the thighs, shanks, and feet were tracked to calculate basic gait features such as average walking speed, average time between steps, and 72 other measures, including their standard deviations. When training our algorithms on 70% of the data, and testing the remaining 30%, we achieved an identification accuracy of 93.21% to 99.38%. When training our algorithms on the first 9 trials and testing the remaining 9 trials, we achieved an identification accuracy of 68.89% to 95.93%. Finally, when training our algorithms on the first trial and testing the remaining 17 trials, we achieved an identification accuracy of 46.27 to 88.43%. Our results provide preliminary evidence that gait could be a distinguishing feature in humans. Optimization efforts are ongoing to reach perfect identification accuracy in the future. This project is supported by NSF 212491, NIH P20GM109090, R01NS114282, University of Nebraska Collaboration Initiative, the Center for Research in Human Movement Variability at the University of Nebraska at Omaha, NASA EPSCoR, IARPA.
LOWER BODY GAIT VARIABILITY AS A DISTINGUISHING FEATURE IN HUMANS
CEC RM #128
Most humans have a fingerprint that is unique and persists throughout life. The same may be true for gait. Each person is unique physiologically and biomechanically, and has experienced a unique progression of life including injuries, habits, hobbies, or lifestyles that all contribute to navigating the world. Those individualized experiences, whether intended or not, are potentially on display in gait. We hypothesized that there are gait characteristics intrinsic and unique to everyone, so that everyone has a unique “gaitprint”, similar to humans possessing unique fingerprints. To test our hypothesis, we recruited thirty healthy young adults between the ages 19-35 as part of the NONAN GaitPrint project. In this study we put simple gait features through several distance and machine learning algorithms to identify people based on how they walk. Each participant completed 18, 4-minute walking trials at a self-selected pace on an indoor track. The motion of the thighs, shanks, and feet were tracked to calculate basic gait features such as average walking speed, average time between steps, and 72 other measures, including their standard deviations. When training our algorithms on 70% of the data, and testing the remaining 30%, we achieved an identification accuracy of 93.21% to 99.38%. When training our algorithms on the first 9 trials and testing the remaining 9 trials, we achieved an identification accuracy of 68.89% to 95.93%. Finally, when training our algorithms on the first trial and testing the remaining 17 trials, we achieved an identification accuracy of 46.27 to 88.43%. Our results provide preliminary evidence that gait could be a distinguishing feature in humans. Optimization efforts are ongoing to reach perfect identification accuracy in the future. This project is supported by NSF 212491, NIH P20GM109090, R01NS114282, University of Nebraska Collaboration Initiative, the Center for Research in Human Movement Variability at the University of Nebraska at Omaha, NASA EPSCoR, IARPA.