Pattern detection of atherosclerosis from carotid artery doppler signals using fuzzy weighted pre-processing and...

Area of Science:

• Medical Imaging
• Biomedical Signal Processing
• Artificial Intelligence in Medicine

Background:

• Atherosclerosis diagnosis relies on interpreting Carotid Artery Doppler Signals.
• Accurate and early detection of Atherosclerosis is crucial for patient outcomes.
• Current diagnostic methods may benefit from advanced signal processing and machine learning.

Purpose of the Study:

• To develop and evaluate an expert system for Atherosclerosis diagnosis using Carotid Artery Doppler Signals.
• To assess the efficacy of Fuzzy weighted pre-processing and Least Square Support Vector Machine (LSSVM) in Atherosclerosis detection.
• To achieve high classification accuracy in distinguishing between healthy controls and Atherosclerosis patients.

Main Methods:

• Recorded Carotid Artery Doppler Signals from 114 subjects (60 with Atherosclerosis, 54 healthy controls).
• Performed spectral analysis using Autoregressive (AR) modeling to determine LSSVM inputs.
• Applied a fuzzy weighted pre-processing expert system to the spectral analysis inputs.
• Utilized Least Square Support Vector Machine (LSSVM) for Atherosclerosis classification.

Main Results:

• The expert system achieved 100% classification accuracy in detecting Atherosclerosis.
• The methodology demonstrated high efficacy in differentiating between healthy and diseased subjects.
• 10-fold Cross Validation (CV) confirmed the robustness of the diagnostic model.

Conclusions:

• The proposed expert system effectively diagnoses Atherosclerosis from Carotid Artery Doppler Signals.
• Fuzzy weighted pre-processing combined with LSSVM offers a highly accurate diagnostic approach.
• This method shows significant potential for improving Atherosclerosis detection in clinical practice.