Design of Hardware Accelerators for Optimized and Quantized Neural Networks to Detect Atrial Fibrillation in Patch ECG Device with RISC-V
Sensors · 2023
DOI: 10.3390/s23052703 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, 2023. Verfasst von Ingo Hoyer und 6 weiteren Autor:innen. Beteiligte Einrichtung: Fraunhofer Institute for Microelectronic Circuits and Systems, 47057 Duisburg, Germany. Hauptthemen laut MEDLINE: Atrial Fibrillation. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Federal Ministry of Education and Research.
Abstract (Originalsprache)
Atrial Fibrillation (AF) is one of the most common heart arrhythmias. It is known to cause up to 15% of all strokes. In current times, modern detection systems for arrhythmias, such as single-use patch electrocardiogram (ECG) devices, have to be energy efficient, small, and affordable. In this work, specialized hardware accelerators were developed. First, an artificial neural network (NN) for the detection of AF was optimized. Special attention was paid to the minimum requirements for the inference on a RISC-V-based microcontroller. Hence, a 32-bit floating-point-based NN was analyzed. To reduce the silicon area needed, the NN was quantized to an 8-bit fixed-point datatype (Q7). Based on this datatype, specialized accelerators were developed. Those accelerators included single-instruction multiple-data (SIMD) hardware as well as accelerators for activation functions such as sigmoid and hyperbolic tangents. To accelerate activation functions that require the e-function as part of their computation (e.g., softmax), an e-function accelerator was implemented in the hardware. To compensate for the losses of quantization, the network was expanded and optimized for run-time and memory requirements. The resulting NN has a 7.5% lower run-time in clock cycles (cc) without the accelerators and 2.2 percentage points (pp) lower accuracy compared to a floating-point-based net, while requiring 65% less memory. With the specialized accelerators, the inference run-time was lowered by 87.2% while the F1-Score decreased by 6.1 pp. Implementing the Q7 accelerators instead of the floating-point unit (FPU), the silicon area needed for the microcontroller in 180 nm-technology is below 1 mm2.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Fraunhofer Institute for Microelectronic Circuits and Systems, 47057 Duisburg, Germany.
- KIS*MED (AI Systems in Medicine), Technical University of Darmstadt, 64289 Darmstadt, Germany.
- Department of Electronic Components and Circuits, University of Duisburg-Essen, 47057 Duisburg, Germany.
ORCID-Profile: Ingo Hoyer, Alexander Utz, André Lüdecke, Maurice Rohr, Christoph Hoog Antink, Karsten Seidl
Förderung
- Federal Ministry of Education and Research
Zitieren
APA
Hoyer, I., Utz, A., Lüdecke, A., Kappert, H., Rohr, M., Antink, C. H., & Seidl, K. (2023). Design of Hardware Accelerators for Optimized and Quantized Neural Networks to Detect Atrial Fibrillation in Patch ECG Device with RISC-V. Sensors. https://doi.org/10.3390/s23052703
Vancouver
Hoyer I, Utz A, Lüdecke A, Kappert H, Rohr M, Antink CH, et al. Design of Hardware Accelerators for Optimized and Quantized Neural Networks to Detect Atrial Fibrillation in Patch ECG Device with RISC-V. Sensors. 2023. doi:10.3390/s23052703
BibTeX
@article{Hoyer2023,
author = {Hoyer, Ingo and Utz, Alexander and Lüdecke, André and Kappert, Holger and Rohr, Maurice and Antink, Christoph Hoog and Seidl, Karsten},
title = {Design of Hardware Accelerators for Optimized and Quantized Neural Networks to Detect Atrial Fibrillation in Patch ECG Device with RISC-V},
journal = {Sensors},
year = {2023},
doi = {10.3390/s23052703}
}
EUVIMED