Verhulst map measures: new biomarkers for heart rate classification

Area of Science:

• Biomedical Engineering
• Nonlinear Dynamics
• Medical Signal Processing

Background:

• Analyzing biological signals is crucial for understanding bio-systems.
• Traditional linear methods fail to capture non-stationary properties of bio-signals.
• Nonlinear dynamical techniques are vital for examining complex bio-system behavior.

Purpose of the Study:

• Propose novel bio-markers derived from the chaotic nature of biomedical signals.
• Utilize the Verhulst map to characterize reconstructed phase space morphology.
• Develop a framework for analyzing and classifying heart rate (HR) signals.

Main Methods:

• Extracted features from heart rate (HR) signals of meditator and non-meditator groups.
• Employed the Verhulst map for bio-marker development.
• Evaluated conventional classifiers (k-NN, SVM, Naïve Bayes) and a hybrid majority voting strategy.

Main Results:

• Achieved a maximum accuracy, F1-score, and sensitivity of 100%.
• Demonstrated the framework's capability in analyzing and classifying HR signals.
• Highlighted the effectiveness of Verhulst diagram-based measures.

Conclusions:

• Verhulst diagram-based measures offer a simple, dynamics-based approach for bio-signal quantification.
• The proposed framework excels in analyzing and classifying heart rate signals.
• This method holds potential for various medical signal analysis applications.