Sleep apnoea detection in single channel ECGs by analyzing heart rate dynamics

C. Zywietz,V. von Einem,B. Widiger,G. Joseph
DOI: https://doi.org/10.1109/IEMBS.2001.1020574
2001-10-25
Abstract:Sleep apnoea is often cause of/or associated with hypertension and of risk factors like "seconds sleep" during the day. Sleep disorders are typically investigated by means of polysomnographic recordings. We have analyzed 70 eight-hour single-channel ECG recordings to rind out to which extent sleep apnoeas may be detected from the ECG alone. From the 70 data sets 35 were annotated by experts for phases of regular sleep and phases with sleep apnoea. Our analysis is based on spectral components of heart rate variability. Frequency analysis was performed using Fourier and wavelet transformation with appropriate application of the Hilbert transform. Classification is based on four frequency bands. We defined: ULF Band (0-0.013 Hz), VLF Band (0.013-0.0375 Hz), LF Band (0.0375-0.06 Hz) and the HF Band (0.17-0.28 Hz). For classification linear discriminant functions were applied using spectral components and other variables derived from the records. Classification was made for patient records as a whole and for the minutes in each of the recordings. Classification of cases was based on three variables. For the Learning Set a sensitivity for apnoea of 95.0% at a specificity of 100% was achieved. For the minutes allocation a sensitivity of 90.8% at a specificity of 92.7% was obtained.
What problem does this paper attempt to address?