Heart rate monitoring and atrial fibrillation detection smartphone apps (iOS)

Abstract

The main objective of this project is to develop two Photoplethysmography (PPG) based cardiac monitoring apps for iOS devices. The first app is "My Heart Rate", which is used for monitoring heart rate before and after physical exercises. As this is the second version of this app, new functionalities, system architecture and user interface (UI) are used to enhance user experience. Automatic fingertip detection is introduced to ensure proper usage and reliable results, with error detection based on the histogram of captured video frame by smartphone’s camera. Mean and variance of the pixels’ intensity in RGB channels are used to identify red frame. The amplitude of PPG signal is analyzed to identify fingertip. The second app is "Photo AFib", which is designed for detecting Atrial Fibrillation (AF) based on the heart rate retrieval and error detection algorithm. The conventional methods to detect this paroxysmal arrhythmia are expensive and inconvenient. The app analyzes the realtime PPG signal with two statistical methods, which are Root Mean Square of Successive Differences (RMSSD) and Shannon Entropy (ShE). The analyzed peak-to-peak variability and complexity are compared with threshold obtained from MIT-BIH AF library. The results are compared with the medical proved AF detecting application.