**Abstract**

Parkinson's disease (PD) is a chronic neurodegenerative disease that affects 1% of the population and whose diagnosis is considered one of the most challenges in the area of neurology. The goal of our work is to assist physicians with the correct diagnosis and early detection of PD. This chapter provides a review of previous work on PD detection under two perspectives, voice analysis and Magnetic Resonance Imaging analysis, by comparing our work with those from other authors. For the case of voice-based PD detection, accuracy reaches 95.9% in female patients and 94.36% in male patients on the largest available dataset. Another contribution in this area is the analysis of voice features to assist the clinical interpretation of the binary result of voice-based detection. For the case of structural Magnetic Resonance Imaging (sMRI)-based PD detection, detection accuracy reaches 96.97% for female patients and 99.01% for male patients using the Parkinson's Progression Marker Initiative dataset. We provide a discussion about the finding of new regions of interest to assist in the detection of PD on sMRI. There is also a comparison between voice-based and MRI-based PD detection methods. Finally, a perspective on future work for PD detection is discussed.

**Keywords:** Parkinson's disease, machine learning, biomedical engineering, magnetic resonance imaging, voice analysis, diagnostic tool
