Long Short-Term Memory for apnea detection based on Heart Rate Variability
- 1 August 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2008 (1094687X) , 5234-5237
Abstract
The main drive force in apnea current diagnostic is to reduce overwhelming number of sleep disorders candidates by means of very simple-to-use, comfortable and cheap methodology. The proposed framework is based only on automatic analysis of electrocardiogram signal. The feature extraction stage was performed using methods of Heart Rate Variability and Detrended Fluctuation analysis. Feature-spaces formed using these two methods were used as input to a Long Short-Term Memory Artificial Neural Network chosen for its capability to find temporally dependencies in the data. The framework was evaluated on Challenge 2000 Physionet database yielding successful rate 82.1%, sensitivity 85.5% and specificity 80.1%.Keywords
This publication has 7 references indexed in Scilit:
- Obstructive sleep apnea classification based on spectrogram patterns in the electrocardiogram2002
- Systematic comparison of different algorithms for apnoea detection based on electrocardiogram recordingsMedical & Biological Engineering & Computing, 2002
- Estimation of parameters and eigenmodes of multivariate autoregressive modelsACM Transactions on Mathematical Software, 2001
- Long Short-Term MemoryNeural Computation, 1997
- Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time seriesChaos: An Interdisciplinary Journal of Nonlinear Science, 1995
- Power spectrum analysis of heart rate variability to assess the changes in sympathovagal balance during graded orthostatic tilt.Circulation, 1994
- A Real-Time QRS Detection AlgorithmIEEE Transactions on Biomedical Engineering, 1985