Key result
A classification system combining Linear Predictive Coding and a Support Vector Machine with a Modified Cuckoo Search optimizer achieved 93.23% accuracy in simultaneously identifying 12 different heart sound classes.
Absolute Event Rate: 93.23% vs 71.83%
A novel heart sound classification system combining Linear Predictive Coding, Support Vector Machine, and Modified Cuckoo Search achieved >93% accuracy across 12 heart sound classes.
No takes yet. Share an insight, caveat, or question.
May aid automated multi-class heart sound analysis; leaves open prospective clinical validation before adoption.
Redlarski et al. (2014) studied Heart sounds classification (n=72). Support Vector Machine with Modified Cuckoo Search (SVM-MCS) classifier vs. Artificial Neural Network and standard SVM classifiers was evaluated on Classification accuracy for 12 simultaneous heart sound classes. A classification system combining Linear Predictive Coding and a Support Vector Machine with a Modified Cuckoo Search optimizer achieved 93.23% accuracy in simultaneously identifying 12 different heart sound classes.
Synapse has enriched one closely related paper. Consider it for comparative context: