Classification of PD patients and controls.

The pattern recognition algorithm “APD” identified an optimal classifier and feature combination to reveal balanced classification rates, sensitivity, specificity, and positive predictive value (PPV) for the experiment PD patients vs. controls using the AdaBoost classifier and cross-validation in the training sample. This algorithm was validated using an independent validation sample. Features used for this algorithm include feature no.: 2, 5, 6, 8, 10, 11, 12, 13, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 derived from the 10-meter walk, heel-toe tapping, and circling test (for detailed description see table S3).