An Artificial Intelligence Olfactory-Based Diagnostic Model for Parkinson's Disease Using Volatile Organic Compounds...

Area of Science:

• Biochemistry
• Neuroscience
• Analytical Chemistry

Background:

• Parkinson's Disease (PD) is a prevalent neurodegenerative disorder requiring early diagnosis.
• Current diagnostic methods for PD can be invasive or lack sensitivity.
• Volatile organic compounds (VOCs) in biological samples offer potential as non-invasive biomarkers.

Purpose of the Study:

• To develop a novel diagnostic model for Parkinson's Disease (PD).
• To identify specific VOC biomarkers in ear canal secretions (ECS) for PD detection.
• To enhance PD diagnostic accuracy using integrated sensor and machine learning technologies.

Main Methods:

• Gas chromatography-mass spectrometry (GC-MS) was used to analyze VOCs in ECS samples from PD and non-PD patients.
• Four VOCs (ethylbenzene, 4-ethyltoluene, pentanal, and 2-pentadecyl-1,3-dioxolane) were identified as significant biomarkers.
• A diagnostic model integrating gas chromatography-surface acoustic wave sensors (GC-SAW) and a convolutional neural network (CNN) was developed.

Main Results:

• Statistically significant differences in VOC profiles were observed between PD and non-PD groups.
• The identified VOC biomarkers demonstrated strong capability in classifying PD patients.
• The GC-SAW-CNN model achieved a diagnostic accuracy of up to 94.4% for PD detection.

Conclusions:

• Specific VOCs in ECS can serve as reliable biomarkers for Parkinson's Disease diagnosis.
• The integrated GC-SAW-CNN approach offers a highly accurate and efficient method for PD detection.
• This technology holds promise for a new, non-invasive bedside diagnostic device for PD.