Convolutional neural networks-based early Parkinson’s disease classification using cycling data from a steerable indoor bicycle
Kim Y, Kim J, Kang S, Lee Y, Moon J, Kim SJ, Kim BJ et al.
CNN model achieved ~86.1% accuracy in early Parkinson’s disease detection using cycling-derived kinetic and kinematic data from the Ultiracer platform.
29 PD patients and 36 healthy controls. Used two 6-axis force-torque sensors (headset spacer and seat post). Input data: 30 seconds of cycling data (force, moment, speed, lateral movement) plus demographic data.



