Cardio-EPNet: Cardiovascular Disease Classification Via Hybrid Dual-Path Edge and Pixel Aware Neural Network

Key Points

• The aim is to develop a hybrid dual-path neural network for better classification of cardiovascular diseases.
• Utilized a hybrid model combining edge and pixel aware features for analysis.
• Employed neural network training on a dataset of cardiovascular images.
• Validated classification accuracy against standard medical diagnoses.
• Achieved a classification accuracy of 92% (95% CI, 90%-94%), p<0.001.
• Demonstrated improved sensitivity for detecting cardiovascular diseases.
• Reduced false positive rates by 15% compared to traditional methods.