Chris Bailey

Classification of male and female walking strategies (University of Ottawa) Sep 2022 – Present

Developing a k-means clustering model in Python to classify nuanced movement strategies from a large array of noisy optoelectronic and wearable sensor time series data
The model output would be used as a feature to better explain an individual’s risk of falling
Awarded $90,000 in federal funding from the Natural Sciences and Engineering Research Council of Canada (NSERC)

Prediction of walking stability using a smartwatch (University of Ottawa) Sep 2021 – Present

Principal Investigator of a successful grant awarded by Apple Inc.
Led a 6-member team in experimental design and in-lab data acquisition from 41 participants using 4 asynchronous motion capture systems
Created programs in Matlab and in Python to synchronize, aggregate, explore, and clean terabytes of structured and semi-structured data containing hundreds of features
Predicted walking stability to within 7% of gold standard by training regression models using support vector machines, extreme gradient boosting, and LSTM networks on features I engineered from an Apple Smartwatch using principal component analysis
The smartwatch alternative to the gold standard ($100,000+ set of location-restrictive hardware) enables an inexpensive option to monitor walking stability anywhere and anytime

Simulation of movement variability from wearable sensors (University of Ottawa) Sep 2020 – Aug 2021

Pioneered and validated the first simulation of trunk and lower limb movement variability in an open-source musculoskeletal model from 8 wearable sensors, using the OpenSim API in Matlab
Optimized parameter weights to enhance simulation accuracy by 25%
The wearable sensor-based simulation measures 18 joint angles (outputs key to understanding movement strategies) to within 5° of gold standard at <1% of gold standard cost
Awarded $90,000 in funding from the Children’s Hospital of Eastern Ontario Research institute

Characterization of movement variability in aging men and women (McGill University) Sep 2016 – Aug 2020

Applied field knowledge and principal components analysis to engineer features from biological time series data (kinematics, electromyography, near-infrared spectroscopy, ultrasonography)
Developed a Matlab program to automatically explore, clean, and process time series data across thousands of files acquired from aging populations, saving my team weeks of time
Debugged and refined numerous Matlab programs for greater user-friendliness by incorporating file selection UIs, creating a batch processing option, and handling different data input formats
Established the currently-used lab data visualization, poster, and presentation formats

About