Template Matching Based Early Exit CNN for Energy-efficient Myocardial Infarction Detection on Low-power Wearable Devices

Template Matching Based Early Exit CNN for Energy-efficient Myocardial Infarction Detection on Low-power Wearable Devices
复制标题

基于模板匹配的早期退出 CNN,用于低功耗可穿戴设备上的节能心肌梗塞检测

DOI:
10.1145/3534580
复制
发表时间:
2022
期刊:
Wearable and Ubiquitous Technologies
影响因子:
--
通讯作者:
Al Faruque, Mohammad Abdullah
Al Faruque, Mohammad Abdullah
中科院分区:
--
文献类型:
--
作者:
Rashid, Nafiul;Demirel, Berken Utku;Odema, Mohanad;Al Faruque, Mohammad Abdullah

文献摘要

参考文献

被引文献

相似文献

心肌梗死(MI),也称为心脏病发作,是一种危及生命的心脏病,是全球死亡的主要原因。其反复出现和无声的性质强调了通过可穿戴设备进行持续监测的必要性。可穿戴设备解决方案应该提供足够的性能,同时在功率和存储器方面受到资源约束。本文提出了一种使用卷积神经网络(CNN)的MI检测方法,该方法在两个数据集(PTB和PTB-XL)的可穿戴设备上的性能优于最先进的工作,同时具有能量和内存效率。此外,我们还提出了一种新的基于模板匹配的早期退出(TMEX)CNN架构,与基线架构相比,该架构进一步提高了能效,同时保持了类似的性能。我们的基线和TMEX架构在PTB数据集上实现了99.33%和99.24%的准确率,而在PTB-XL数据集上,它们分别实现了84.36%和84.24%的准确率。这两种架构都适用于仅需要20 KB RAM的可穿戴设备。对真实的硬件的评估表明,我们的基准架构比可穿戴设备上的最先进的工作高出0.6倍至53倍的能效。此外,我们的TMEX架构进一步提高了8.12%(PTB)和6.36%(PTB-XL)的能源效率,同时保持与基线架构相似的性能。
Myocardial Infarction (MI), also known as heart attack, is a life-threatening form of heart disease that is a leading cause of death worldwide. Its recurrent and silent nature emphasizes the need for continuous monitoring through wearable devices. The wearable device solutions should provide adequate performance while being resource-constrained in terms of power and memory. This paper proposes an MI detection methodology using a Convolutional Neural Network (CNN) that outperforms the state-of-the-art works on wearable devices for two datasets - PTB and PTB-XL, while being energy and memory-efficient. Moreover, we also propose a novel Template Matching based Early Exit (TMEX) CNN architecture that further increases the energy efficiency compared to baseline architecture while maintaining similar performance. Our baseline and TMEX architecture achieve 99.33% and 99.24% accuracy on PTB dataset, whereas on PTB-XL dataset they achieve 84.36% and 84.24% accuracy, respectively. Both architectures are suitable for wearable devices requiring only 20 KB of RAM. Evaluation of real hardware shows that our baseline architecture is 0.6x to 53x more energy-efficient than the state-of-the-art works on wearable devices. Moreover, our TMEX architecture further improves the energy efficiency by 8.12% (PTB) and 6.36% (PTB-XL) while maintaining similar performance as the baseline architecture.
私有和安全领域的自我感知雾计算
DOI: 10.1007/978-3-319-57639-8_5
发表时间: 2018
期刊: 2019 10th International Conference on Information, Intelligence, Systems and Applications (IISA)
影响因子: --
作者:
Kalle Tammemäe;A. Jantsch;A. Kuusik;J. Preden;E. Ounapuu
通讯作者: E. Ounapuu
DOI: 10.1109/embc44109.2020.9175232
发表时间: 2020
期刊: 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC)
影响因子: --
作者:
Nafiul Rashid;M. A. Faruque
通讯作者: M. A. Faruque
DOI: --
发表时间: 2019-05
期刊: --
影响因子: --
作者:
Xuanyi Dong;Yi Yang
通讯作者: Xuanyi Dong;Yi Yang
用于身体和情绪健康监测的可穿戴设备
DOI: 10.1109/cic.2015.7408601
发表时间: 2015
期刊: 2015 Computing in Cardiology Conference (CinC)
影响因子: --
作者:
S. Murali;F. Rincón;David Atienza Alonso
通讯作者: David Atienza Alonso
TEA-DNN:寻求时间-能量-精度协同优化的深度神经网络
DOI: 10.1109/islped.2019.8824934
发表时间: 2018
期刊: 2019 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED)
影响因子: --
作者:
Lile Cai;Anne;Arthur Herbout;Chuan;Jie Lin;V. Chandrasekhar;M. Sabry
通讯作者: M. Sabry