Prediabetes risk classification algorithm via carotid bodies and K-means clustering technique

Area of Science:

• Biomedical Engineering
• Machine Learning
• Cardiovascular Physiology

Background:

• Type-2 diabetes mellitus (T2DM) affects millions globally, necessitating early detection to prevent complications and reduce healthcare burdens.
• Carotid bodies (CBs) play a role in cardiorespiratory regulation, and their chemosensitivity may be altered in early stages of metabolic dysfunction.
• Current diagnostic methods for prediabetes can be improved with more sensitive and accessible screening tools.

Purpose of the Study:

• To introduce a novel prediabetes risk classification algorithm (PRCA).
• To investigate the utility of carotid body (CB) chemosensitivity in detecting prediabetes.
• To apply machine learning techniques for enhanced diagnostic accuracy.

Main Methods:

• Collected heart rate (HR) and respiratory rate (RR) data from individuals with and without prediabetes.
• Stimulated CB chemosensitivity via oxygen inhalation and assessed responses after a standardized meal.
• Utilized K-means clustering and developed a novel algorithm (PRCA) for risk classification.

Main Results:

• Observed greater cardiorespiratory variability in the prediabetes group compared to controls, particularly after oxygen stimulation.
• The developed PRCA demonstrated 86% accuracy in classifying individuals for prediabetes.
• The findings highlight CB chemosensitivity deregulation as a potential biomarker for early T2DM stages.

Conclusions:

• The PRCA offers a promising, nuanced method for early prediabetes detection based on CB chemosensitivity.
• The machine learning approach and algorithm adaptability suggest potential applications for classifying risks of other diseases.
• Further validation via longitudinal studies and larger cohorts is recommended to confirm the algorithm's efficacy.