Detecting Vascular Age Using the Analysis of Peripheral Pulse.

Area of Science:

• Biomedical Engineering
• Cardiovascular Research
• Computational Biology

Background:

• Vascular aging is linked to arterial stiffening and endothelial dysfunction, increasing cardiovascular disease risk.
• Microvascular pulse analysis is a focus for non-invasive assessment of circulatory system biomechanics.

Purpose of the Study:

• To propose and validate a support vector machine (SVM) learning approach for assessing vascular aging using pulse contour features.
• To develop a non-invasive method for detecting age-related vascular modifications.

Main Methods:

• A multi-Gaussian decomposition algorithm was used to generate 20,935 pulse wave models from laser Doppler flowmetry signals of 54 healthy subjects.
• A supervised support vector machine (SVM) classifier was trained and validated on these pulse wave models.

Main Results:

• The multi-Gaussian model achieved a mean R² of 0.98, accurately reconstructing pulse shapes.
• The SVM demonstrated a mean Pearson's r of 0.808 between classified wave age and actual subject age, with an average ROC AUC of 0.953.

Conclusions:

• The SVM-based method effectively discriminates between individuals of different ages based on vascular aging markers.
• This approach may detect vascular aging and is potentially applicable for screening diabetic angiopathy due to similar vascular alterations.