A novel multi-headed Conv-LSTM network for heart rate estimation using wrist-worn PPG and acceleration signals achieved an average mean absolute error of 6.28 bpm and a Pearson's correlation of 0.85.
A novel MH Conv-LSTM DeepPPG model improves heart rate estimation accuracy from wrist-worn wearable devices during daily activities.
Effect estimate: Pearson's correlation 0.85
Non-invasive photoplethysmography (PPG) technology was developed to track heart rate during physical activity under free-living conditions. Automated analysis of PPG has made it useful in both clinical and non-clinical applications. Because of their generalization capabilities, deep learning methods can be a major direction in the search for a heart rate estimation solution based on signals from wearable devices. A novel multi-headed convolutional neural network model enriched with long short-term memory cells (MH Conv-LSTM DeepPPG) was proposed for the estimation of heart rate based on signals measured by a wrist-worn wearable device, such as PPG and acceleration signals. For the PPG-DaLiA dataset, the proposed solution improves the performance of previously proposed methods. An experimental approach was used to develop the final network architecture. The average mean absolute error (MAE) of the final solution was 6.28 bpm and Pearson's correlation coefficient between the estimated and true heart rate values was 0.85.
Wilkosz et al. (Sat,) conducted a other in Heart rate estimation during daily living activities. MH Conv-LSTM DeepPPG model vs. Previously proposed methods was evaluated on Heart rate estimation accuracy (Mean absolute error and Pearson's correlation) (Pearson's correlation 0.85). A novel multi-headed Conv-LSTM network for heart rate estimation using wrist-worn PPG and acceleration signals achieved an average mean absolute error of 6.28 bpm and a Pearson's correlation of 0.85.