Dynamic learning of heart sounds with changing noise
Yiqi Deng, Peter J. Bentley · 2014
Agent-Based Models are used to model dynamic systems such as stock markets, societies, and complex biological systems that are difficult to model analytically using partial differential equations. Many agent-based modeling software are designed for serial von-Neumann computer architectures. That limits the speed and scalability of these systems. Systemic computation (SC) is designed to be a model of natural behavior and, at the same time, a non Von-Neumann architecture with its characteristics similar to multi-agent system. Here we propose a novel method based on an Artificial Immune System (AIS) and implemented on a systemic computer, which is designed to adapt itself over continuous arrival of data to cope with changing patterns of noise without requirement for feedback, as a result of its own experience. Experiments with heartbeat data collected from a clinical trial in hospitals using a digital stethoscope shows the algorithm performs up to 3.60% better in the precision rate of murmur and 3.96% better in the recall rate of murmur than other standard anomaly detector approaches such as Multiple Kernel Anomaly Detection (MKAD).