TY - GEN
T1 - Recurrent neural network circuit for automated detection of atrial fibrillation from raw ECG
AU - Sadasivuni, Sudarsan
AU - Chowdhury, Rahul
AU - Karnam, Vinay Elkoori Ghantala
AU - Banerjee, Imon
AU - Sanyal, Arindam
N1 - Funding Information:
ACKNOWLEDGMENT This material is based on research sponsored by Air Force Research Laboratory under agreement number FA8650-18-2-5402. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of Air Force Research Laboratory or the U.S. Government.
Publisher Copyright:
© 2021 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
PY - 2021
Y1 - 2021
N2 - A recurrent neural network (RNN) is presented in this work for automatic detection of atrial fibrillation from raw ECG signals without any hand-crafted feature extraction. We designed a stacked long-short term memory (LSTM) network - a special RNN with capability of learning long-term temporal dependencies in the ECG signal. The RNN is digitally synthesized in 65nm CMOS process, and consumes 21.8nJ/inference at 1kHz operating frequency, while achieving state-of-the-art classification accuracy of 85.7% and f1-score of 0.82. The energy consumption of the proposed RNN is 8× lower than state-of-the-art integrated circuits for arrhythmia detection.
AB - A recurrent neural network (RNN) is presented in this work for automatic detection of atrial fibrillation from raw ECG signals without any hand-crafted feature extraction. We designed a stacked long-short term memory (LSTM) network - a special RNN with capability of learning long-term temporal dependencies in the ECG signal. The RNN is digitally synthesized in 65nm CMOS process, and consumes 21.8nJ/inference at 1kHz operating frequency, while achieving state-of-the-art classification accuracy of 85.7% and f1-score of 0.82. The energy consumption of the proposed RNN is 8× lower than state-of-the-art integrated circuits for arrhythmia detection.
KW - Atrial fibrillation
KW - Electro-cardiograph
KW - Health monitoring
KW - Long-short term memory
KW - Recurrent neural network
UR - http://www.scopus.com/inward/record.url?scp=85109044103&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85109044103&partnerID=8YFLogxK
U2 - 10.1109/ISCAS51556.2021.9401666
DO - 10.1109/ISCAS51556.2021.9401666
M3 - Conference contribution
AN - SCOPUS:85109044103
T3 - Proceedings - IEEE International Symposium on Circuits and Systems
BT - 2021 IEEE International Symposium on Circuits and Systems, ISCAS 2021 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 53rd IEEE International Symposium on Circuits and Systems, ISCAS 2021
Y2 - 22 May 2021 through 28 May 2021
ER -