ECG Arrhythmia Classification on an Ultra-Low-Power Microcontroller

Dekimpe, Rémi; Bol, David

doi:10.1109/tbcas.2022.3182159

Cited by 15 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We would like to stress that the proposed ECG simulator can generate a wide range of pathological conditions, so it provides a promising platform not only to train arrhythmia classifiers and detectors [ 63 , 64 ] but also to assess ECG signal processing software. Signal denoising is a remarkable example where synthetic waveforms can help attenuate the noise of real ECG signals by providing a reference profile to wave detector [ 65 ].…”

Section: Resultsmentioning

confidence: 99%

ECG Patient Simulator Based on Mathematical Models

Quiroz-Juárez

Rosales-Juárez

Jiménez-Ramírez

et al. 2022

Sensors

View full text Add to dashboard Cite

In this work, we propose a versatile, low-cost, and tunable electronic device to generate realistic electrocardiogram (ECG) waveforms, capable of simulating ECG of patients within a wide range of possibilities. A visual analysis of the clinical ECG register provides the cardiologist with vital physiological information to determine the patient’s heart condition. Because of its clinical significance, there is a strong interest in algorithms and medical ECG measuring devices that acquire, preserve, and process ECG recordings with high fidelity. Bearing this in mind, the proposed electronic device is based on four different mathematical models describing macroscopic heartbeat dynamics with ordinary differential equations. Firstly, we produce full 12-lead ECG profiles by implementing a model comprising a network of heterogeneous oscillators. Then, we implement a discretized reaction–diffusion model in our electronic device to reproduce ECG waveforms from various rhythm disorders. Finally, in order to show the versatility and capabilities of our system, we include two additional models, a ring of three coupled oscillators and a model based on a quasiperiodic motion, which can reproduce a wide range of pathological conditions. With this, the proposed device can reproduce around thirty-two cardiac rhythms with the possibility of exploring different parameter values to simulate new arrhythmias with the same hardware. Our system, which is a hybrid analog–digital circuit, generates realistic ECG signals through digital-to-analog converters whose amplitudes and waveforms are controlled through an interactive and friendly graphic interface. Our ECG patient simulator arises as a promising platform for assessing the performance of electrocardiograph equipment and ECG signal processing software in clinical trials. Additionally the produced 12-lead profiles can be tested in patient monitoring systems.

show abstract

Section: Resultsmentioning

confidence: 99%

ECG Patient Simulator Based on Mathematical Models

Quiroz-Juárez

Rosales-Juárez

Jiménez-Ramírez

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Optimization of CNNs in ML and DL commonly involves pruning and quantization. The efficiency of the SCA pruning method in SVM models can be successfully implemented on a Cortex-M4 microcontroller [22]. Effective parameter pruning and L1-norm feature map pruning exemplified to substantially reduce model size in 1D-CNN and can even achieve a balance between reduced power consumption and high accuracy on the ZYNQ Ultrascale FPGA [13,23,24].…”

Section: B Optimizing Cnns For Ectopic Beat Detection Through Pruningmentioning

confidence: 99%

Design and Implementation of an ARM-Based AI Module for Ectopic Beat Classification Using Custom and Structural Pruned Lightweight CNN

Xie,

Lin,

Lee

et al. 2024

IEEE Sensors J.

View full text Add to dashboard Cite

This paper presents the development and testing of lightweight and power-efficient CNN models for ectopic beat classification, tailored for a compact-sized (25x45 mm) ARM-based (STM32H7) AI module. Methods: Two custom lightweight architectures (LMUEBCNet and SEmbedNet) were introduced, and their performances benchmarked against conventional models (AlexNet and VGG19). A structural pruning method, filter pruning via the Taylor score, was employed for pre-trained models' parameter optimizations. Further resizing the input image size to pixel-56/112 for the efficiency of the embedded system. Validation was conducted utilizing a combined ECG simulator and the PhysioNet MIT-BIH arrhythmia ECG dataset, adhering to the ANSI/AAMI EC57 standard. Results: SEmbedNet with parameter size of 0.04M and LMUEBCNet with 0.15M parameters reported accuracy of 99.9%/97.6%. In comparison, the pruned models of VGG19 and AlexNet, which had parameter counts of 0.10M and 0.07M respectively, achieved accuracy of 96.7%/94.4%. Post-structural pruning, VGG19 and AlexNet models saw reductions in parameters by thousands of times, with accuracy decreases of 2% to 5%. Notably, the proposed LMUEBCNet and SEmbedNet were vastly leaner by 930 and 320 times than VGG19 and AlexNet. Conclusion: The proposed ARM-based AI module integrated with custom lightweight CNN models offers superior accuracy-memory trade-offs with 0.04M parameters, which is less than 1% the size of conventional models. The AI module with compact size and a power consumption of only 0.4 watts achieves a classification rate of 4.2 segments per second and 99.9% accuracy. The AI module shows the potential to transform ECG monitoring devices into ECG analyzers.

show abstract

“…[95] A recent design, drawing more current, that is, 2.5 μA, effectively suppressed the noise to 40 nV/√Hz (0.516 μV rms ) within a compact area of 0.3 mm 2 . [96] In this view, Rémi Dekimpe et al reported the chopping stabilized capacitive coupled instrument amplifier (CS-CCIA) to achieve 750 nV rms IRN at 1.17 μA current and 0.024 mm 2 area, [97] showing excited noise effective factor (NEF) and area efficiency. Recently, a group chopping approach was devised for an 8-channel AFE, obtaining a remarkable 420 nV rms IRN at a minimum cost of 2.4 μW power and 0.017 mm 2 area.…”

Section: Noise Suppressionmentioning

confidence: 99%

High‐density implantable neural electrodes and chips for massive neural recordings

Wang,

Suo,

Wang

et al. 2024

Brain-X

View full text Add to dashboard Cite

High‐density neural recordings with superior spatiotemporal resolution powerfully unveil cellular‐scale neural communication, showing great promise in neural science, translational medicine, and clinical applications. To achieve such, many design and fabrication innovations enhanced the electrode, chip, or both for biocompatibility improvement, electrical performance upgrade, and size miniaturization, offering several thousands of recording sites. However, an enormous gap exists along the trajectory toward billions of recording sites for brain scale resolution, posing many more design challenges. This review tries to find possible insight into mitigating the gap by discussing the latest progress in high‐density electrodes and chips for neural recordings. It emphasizes the design, fabrication, bonding techniques, and in vivo performance optimization of high‐density electrodes. It discusses the promising opportunities for circuit‐level and architecture‐level multi‐channel chip design innovations. We expect that joint effort and close collaboration between high‐density electrodes and chips will pave the way to high‐resolution neural recording tools supporting cutting‐edge neuroscience discoveries and applications.

show abstract

ECG Arrhythmia Classification on an Ultra-Low-Power Microcontroller

Cited by 15 publications

References 36 publications

ECG Patient Simulator Based on Mathematical Models

ECG Patient Simulator Based on Mathematical Models

Design and Implementation of an ARM-Based AI Module for Ectopic Beat Classification Using Custom and Structural Pruned Lightweight CNN

High‐density implantable neural electrodes and chips for massive neural recordings

Contact Info

Product

Resources

About