Screening of Non-overlapping Apnea and Non-apnea from Single Lead ECG-apnea Recordings using Time-Frequency Approach

This study focused on extracting to finding differences between apnea events and non-apnea events using time-frequency approach. This approach is of particular relevance to obtain the efficiency and accuracy of the support system for the classification model. Heart rate variability(HRV) was calculated using the statistic and frequency approach based on the time-frequency domain. The analysis of HRV, about the occurrence of the short recording, was performed selecting two segments: a class of apnea events and a class of non-apnea events. The experiment findings of the statistical analysis of our feature extraction showed time-domain feature estimation with Heart rate means (BPM) slightly higher for non-apnea events about mean ± standard deviation (72(±4)). The frequency-domain features, at VLF, LF and HF power of apnea events, are monitored over time with non-apnea events. The overall experiment indicates a significantly different feature value in the heart rate during examining apnea events and non-apnea events.