A Caltech Library Service

Energy-Efficient Classification for Resource-Constrained Biomedical Applications

Shoaran, Mahsa and Haghi, Benyamin Allahgholizadeh and Taghavi, Milad and Farivar, Masoud and Emami, Azita (2018) Energy-Efficient Classification for Resource-Constrained Biomedical Applications. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 8 (4). pp. 693-707. ISSN 2156-3357. doi:10.1109/JETCAS.2018.2844733.

Full text is not posted in this repository. Consult Related URLs below.

Use this Persistent URL to link to this item:


Biomedical applications often require classifiers that are both accurate and cheap to implement. Today, deep neural networks achieve the state-of-the-art accuracy in most learning tasks that involve large data sets of unstructured data. However, the application of deep learning techniques may not be beneficial in problems with limited training sets and computational resources, or under domain-specific test time constraints. Among other algorithms, ensembles of decision trees, particularly the gradient boosted models have recently been very successful in machine learning competitions. Here, we propose an efficient hardware architecture to implement gradient boosted trees in applications under stringent power, area, and delay constraints, such as medical devices. Specifically, we introduce the concepts of asynchronous tree operation and sequential feature extraction to achieve an unprecedented energy and area efficiency. The proposed architecture is evaluated in automated seizure detection for epilepsy, using 3074 h of intracranial EEG data from 26 patients with 393 seizures. Average F1 scores of 99.23% and 87.86% are achieved for random and block-wise splitting of data into train/test sets, respectively, with an average detection latency of 1.1 s. The proposed classifier is fabricated in a 65-nm TSMC process, consuming 41.2 nJ/class in a total area of 540×1850 μm^2 . This design improves the state-of-the-art by 27× reduction in energy-area-latency product. Moreover, the proposed gradient-boosting architecture offers the flexibility to accommodate variable tree counts specific to each patient, to trade the predictive accuracy with energy. This patient-specific and energy-quality scalable classifier holds great promise for low-power sensor data classification in biomedical applications.

Item Type:Article
Related URLs:
URLURL TypeDescription
Shoaran, Mahsa0000-0002-6426-4799
Haghi, Benyamin Allahgholizadeh0000-0002-4839-7647
Taghavi, Milad0000-0002-5512-0603
Emami, Azita0000-0003-2608-9691
Additional Information:© 2018 IEEE. Manuscript received December 31, 2017; revised May 12, 2018; accepted June 1, 2018. Date of publication June 7, 2018; date of current version December 11, 2018. This work was supported in part by the Heritage Medical Research Institute and in part by the Swiss NSF Fellowship under Grant P300P2-171220.
Group:Heritage Medical Research Institute
Funding AgencyGrant Number
Heritage Medical Research InstituteUNSPECIFIED
Swiss National Science Foundation (SNSF)P300P2-171220
Subject Keywords:Gradient boosted trees, hardware architecture, on-chip classifier, decision tree, accuracy, feature extraction, latency, seizure detection, energy-quality scaling
Issue or Number:4
Record Number:CaltechAUTHORS:20190109-161530924
Persistent URL:
Official Citation:M. Shoaran, B. A. Haghi, M. Taghavi, M. Farivar and A. Emami-Neyestanak, "Energy-Efficient Classification for Resource-Constrained Biomedical Applications," in IEEE Journal on Emerging and Selected Topics in Circuits and Systems, vol. 8, no. 4, pp. 693-707, Dec. 2018. doi: 10.1109/JETCAS.2018.2844733
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:92186
Deposited By: Tony Diaz
Deposited On:10 Jan 2019 12:30
Last Modified:16 Nov 2021 03:47

Repository Staff Only: item control page