Efficient syncope prediction from resting state clinical data using wavelet bispectrum and multilayer perceptron...

Area of Science:

• Cardiology
• Biomedical Engineering
• Physiology

Background:

• Neurally mediated syncope (NMS) is the most common cause of syncope.
• The head up tilt test (HUTT) is the standard diagnostic tool for NMS.
• Predicting NMS before HUTT can improve patient management and reduce healthcare costs.

Purpose of the Study:

• To develop a method for predicting NMS before HUTT using resting physiological data.
• To analyze heart rate variability (HRV) and wavelet higher-order spectrum (WHOS) features for NMS prediction.
• To assess the efficacy of a multilayer perceptron neural network (MPNN) for early NMS detection.

Main Methods:

• Collected HRV data at rest and during HUTT from 26 NMS patients and 10 healthy controls.
• Analyzed time and frequency domain HRV features, including regularity and complexity.
• Utilized WHOS analysis in low-frequency (LF) and high-frequency (HF) bands, combined with systolic blood pressure.
• Trained and validated an MPNN model using 5-fold cross-validation.

Main Results:

• Significant differences in resting HRV entropy and LF-band WHOS features were observed between NMS patients and healthy controls.
• The MPNN model, using resting WHOS-based HRV features (LF band) and systolic blood pressure, achieved 89.7% accuracy.
• These features effectively differentiated between HUTT-positive and HUTT-negative individuals.

Conclusions:

• Resting HRV analysis and systolic blood pressure can predict NMS outcomes prior to HUTT.
• The proposed MPNN model offers a promising approach for early NMS risk identification.
• This method may help identify patients at higher risk for NMS, optimizing diagnostic pathways.