Inflection point analysis: A machine learning approach for extraction of IEGM active intervals during atrial fibrillation
作者：
Highlights：
• Introducing a novel 3D feature space based on characteristics of inflection points of IEGMs.
• Automatic clustering of active and inactive intervals of IEGM using an Expectation Maximization algorithm.
• Higher resolution in estimating the onsets and offsets of IEGMs due to using a non-windowing technique.
• Robustness to noise and baseline variations due to using a Laplacian of Gaussian filter.
• Low computational time which is independent of IEGM sampling rate.
摘要
•Introducing a novel 3D feature space based on characteristics of inflection points of IEGMs.•Automatic clustering of active and inactive intervals of IEGM using an Expectation Maximization algorithm.•Higher resolution in estimating the onsets and offsets of IEGMs due to using a non-windowing technique.•Robustness to noise and baseline variations due to using a Laplacian of Gaussian filter.•Low computational time which is independent of IEGM sampling rate.
论文关键词：Intra-cardiac electrogram,Atrial fibrillation,Inflection point analysis,Gaussian mixture model,Expectation Maximization
论文评审过程：Received 25 September 2017, Revised 11 January 2018, Accepted 15 February 2018, Available online 16 March 2018, Version of Record 16 March 2018.
论文官网地址：https://doi.org/10.1016/j.artmed.2018.02.003
