ECG-SpO2-OSA-Detecting-Obstructive-Sleep-Apnea-using-ECG-and-SpO2-Signals-with-Machine-Learning:

Obstructive sleep apnea (OSA) is a disorder in which a person faces difficulty breathing during their sleep, resulting in severe daytime drowsiness, fatigue, and irritability. Creating an algorithm to assess sleep apnea using the electrocardiogram (ECG) and blood oxygen saturation is the goal of this study (SpO2).

Block Diagram:

Initially, MATLAB was used to find the Wavelet transform, r-r Peaks, RLS filter algorithm, and RR intervals. These algorithms were applied to the dataset, and based on the outcomes it was concluded whether the subject has Sleep Apnea or not.

Machine Learning models used in the project are:

1. Logistic Regression
2. Linear Regression
3. XG Boost
4. K - Means Clustering

MATLAB Output Snapshots:

ML Models Accuracy:

Conclusion:

In this study, a sleep apnea detection method was developed to identify sleep apnea events using ECG and SpO2 datasets. The dataset by physio.net makes full use of the complementary information of the two signals. Thus the method is more accurate than many existing models since SpO2 alone or ECG alone can be used as a potential diagnostic means of SA, but not as a reliable means.