Driving-inspired motor-skill assessment system
A full measurement-and-analysis stack for quantifying a learned motor skill and its decline under amyloid-β pathology.
Stage: active
This is my PhD project: an end-to-end system for measuring a learned motor skill and how it changes under amyloid-β pathology — built from the sensor up.
What it involves
- Custom behavioral apparatus — a driving-inspired task with the instrumentation designed and built for the study.
- Timestamped acquisition — tactile and contact-sensor data captured at roughly 3.6 Hz, alongside 60 fps video.
- Markerless kinematics — pose estimation from the video using DeepLabCut, so movement is quantified without physical markers.
- A reproducible Python pipeline — linear mixed-effects models of the movement trajectories, Mann–Whitney effect sizes, and subject-level aggregation, with explicit data-retention and quality-control reporting across an 84-trial dataset.
Status
The manuscript is in preparation. The analysis code will be released as open source once the work is published — it is not public yet, and this page will link to it when it is.