Why the study?
This study aimed to develop a photoplethysmography-based atrial fibrillation detection algorithm and a method to port it to edge devices.
Does a quantized Transformer-based Deep Neural Network algorithm ported to an edge device improve atrial fibrillation detection accuracy using PPG signals compared to CNN-based and Autoencoder-based models?
Does a quantized Transformer-based Deep Neural Network algorithm ported to an edge device improve atrial fibrillation detection accuracy using PPG signals compared to CNN-based and Autoencoder-based models?
Answers are grounded in this paper and cite their sources.
A quantized Transformer-based deep neural network for atrial fibrillation detection from PPG signals can be successfully compressed to 107 KB and ported to an edge microcontroller while maintaining a high detection accuracy of 98.09%.
This paper proposes an atrial fibrillation detection algorithm using PPG signals and presents a method to port it to edge devices such as the nRF52840. To classify the PPG signals, simple CNN-based, Autoencoder-based, and Transformer-based Deep Neural Networks (DNNs) were implemented and comparatively analyzed using the MIMIC PERform AF dataset. As a result, the Transformer-based algorithm exhibited the best performance with an average accuracy of 98.66%. While post-training quantization (PTQ) of the Transformer-based algorithm caused the average accuracy to degrade to 95.23%, applying Quantization-Aware Training (QAT) prevented this performance drop. Because some layers of the Transformer are not supported in TensorFlow Lite for Microcontrollers (TFLite Micro), they were modified, and the Multi-Head Attention was approximated using a CNN-based network for porting to the edge device, resulting in a final model size of 107 KB. The ported network demonstrated excellent performance, maintaining an average accuracy of 98.09%.
No takes yet. Share an insight, caveat, or question.
Kyoungtaek Choi (2026) studied Atrial fibrillation. Transformer-based Deep Neural Network (DNN) algorithm with Quantization-Aware Training (QAT) ported to edge devices vs. Simple CNN-based and Autoencoder-based Deep Neural Networks (DNNs) was evaluated on Average accuracy of atrial fibrillation detection from PPG signals.
Synapse has enriched 10 related papers. Consider them for comparative context: