東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Analysis and Design of an Energy-Eff...

Miaolin, Zhang,

Linked to FindBook

Google Book

Amazon

博客來

Analysis and Design of an Energy-Efficient Patient-Kspecific Epilepsy Tracking Soc /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Analysis and Design of an Energy-Efficient Patient-Kspecific Epilepsy Tracking Soc // Zhang Miaolin.
Author:	Miaolin, Zhang,
Description:	1 electronic resource (120 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-06, Section: B.
Contained By:	Dissertations Abstracts International86-06B.
Subject:	Electroencephalography. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31777236
ISBN:	9798346773238

Analysis and Design of an Energy-Efficient Patient-Kspecific Epilepsy Tracking Soc /
Miaolin, Zhang,

Analysis and Design of an Energy-Efficient Patient-Kspecific Epilepsy Tracking Soc /Zhang Miaolin. - 1 electronic resource (120 pages)

Source: Dissertations Abstracts International, Volume: 86-06, Section: B.

Epilepsy is a chronic non-communicable disease that affects over 50 million people's lives around the world. Machine learning algorithm utilizing EEG recording to achieve patient-specific real-time ambulatory epilepsy tracking helps the diagnosis and treatment. However, there are some challenges that state-of-the-art epilepsy tracking systems need to overcome: 1) Limited training dataset (for patient-specific training of the classifier), 2) EEG channel selection, 3) transient and localized noise. In this work, we propose a Time-Channel Averaging Feature Extraction algorithm that averages out transient and localized noise, reducing false negative and false positive by 26% and 65%, respectively. Compared with the prevalent minimum redundancy maximum relevance feature selection algorithm, the proposed Time-Averaging based Channel Selection algorithm selects EEG channels with higher average sensitivity and specificity of 1.6%P and 0.4%P. The SoC is implemented in the TSMC 1P8M 40nm process with an area of 1.6mm x 1.3mm. Verified with the CHB-MIT database with more than 180 seizure events and 990 hours of EEG recording of 24 patients, the implemented Guided Time Channel Averaging Support-Vector-Machine classifier achieves sensitivity and specificity of 97.8% and 99.5% across 24 patients 0.68 µJ/Class. The proposed Approximate Online Tuning algorithm uses 512x fewer feature vectors and 874x fewer training clock cycles per tuning compared with previous algorithms. In addition, the online tuning is verified with a local recruited epilepsy patient and achieves a sensitivity boost of 1.8x after one-shot learning and follow-up online tuning for the first time.

English

ISBN: 9798346773238Subjects--Topical Terms:

543293
Electroencephalography.

Analysis and Design of an Energy-Efficient Patient-Kspecific Epilepsy Tracking Soc /
LDR:02950nmm a22003973i 4500 001 2432150
005 20250626110616.5
006 m o d
007 cr|nu||||||||
008 261011s2021 miu||||||m |||||||eng d
020 $a 9798346773238
035 $a (MiAaPQD)AAI31777236
035 $a AAI31777236
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Miaolin, Zhang, $e author. $3 3821283
245 1 0 $a Analysis and Design of an Energy-Efficient Patient-Kspecific Epilepsy Tracking Soc / $c Zhang Miaolin.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 1 electronic resource (120 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-06, Section: B.
500 $a Advisors: Yoo, Jerald.
502 $b Ph.D. $c National University of Singapore (Singapore) $d 2021.
520 $a Epilepsy is a chronic non-communicable disease that affects over 50 million people's lives around the world. Machine learning algorithm utilizing EEG recording to achieve patient-specific real-time ambulatory epilepsy tracking helps the diagnosis and treatment. However, there are some challenges that state-of-the-art epilepsy tracking systems need to overcome: 1) Limited training dataset (for patient-specific training of the classifier), 2) EEG channel selection, 3) transient and localized noise. In this work, we propose a Time-Channel Averaging Feature Extraction algorithm that averages out transient and localized noise, reducing false negative and false positive by 26% and 65%, respectively. Compared with the prevalent minimum redundancy maximum relevance feature selection algorithm, the proposed Time-Averaging based Channel Selection algorithm selects EEG channels with higher average sensitivity and specificity of 1.6%P and 0.4%P. The SoC is implemented in the TSMC 1P8M 40nm process with an area of 1.6mm x 1.3mm. Verified with the CHB-MIT database with more than 180 seizure events and 990 hours of EEG recording of 24 patients, the implemented Guided Time Channel Averaging Support-Vector-Machine classifier achieves sensitivity and specificity of 97.8% and 99.5% across 24 patients 0.68 µJ/Class. The proposed Approximate Online Tuning algorithm uses 512x fewer feature vectors and 874x fewer training clock cycles per tuning compared with previous algorithms. In addition, the online tuning is verified with a local recruited epilepsy patient and achieves a sensitivity boost of 1.8x after one-shot learning and follow-up online tuning for the first time.
546 $a English
590 $a School code: 1883
650 4 $a Electroencephalography. $3 543293
650 4 $a Energy consumption. $3 631630
650 4 $a Support vector machines. $3 2058743
650 4 $a Computer science. $3 523869
650 4 $a Energy. $3 876794
650 4 $a Information technology. $3 532993
650 4 $a Neurosciences. $3 588700
690 $a 0800
690 $a 0984
690 $a 0791
690 $a 0489
690 $a 0317
710 2 $a National University of Singapore (Singapore). $e degree granting institution. $3 3765631
720 1 $a Yoo, Jerald $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-06B.
790 $a 1883
791 $a Ph.D.
792 $a 2021
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31777236