Parkinson’s disease manifests with motor impairments that are detectable through digitized spiral drawings. This study introduces an explainable framework for Parkinson’s disease screening using a novel radial-sampling feature fusion approach. We transform two-dimensional spiral images into one-dimensional revolution signals via a systematic ray-sampling technique to extract three distinct revolutions. We integrate spatial metrics, such as inter-revolution spacing variability and root mean square radial derivatives, with spectral descriptors derived from fast Fourier transform analysis across low-, mid-, and high-harmonic bands. A total of 20 features were utilized to train state-of-the-art machine learning models, including support vector machines, random forests, and light gradient boosting machines. Among these, the random-forest classifier demonstrated superior performance. Subsequent five-fold cross-validation stability analysis, along with feature importance analysis, identified the root mean square radial derivative of the outer revolution (r3_derive_rms) as the most critical biomarker. Stratified cross-validation demonstrates that combining spatial and frequency features significantly enhances detection accuracy compared to single-domain methods, facilitating effective clinical deployment even in data-scarce environments. This interpretable pipeline provides a robust, low-cost “white-box” screening tool, offering a practical alternative to opaque deep-learning models for early clinical intervention.