A Deep Learning Framework for Noise Component Detection from Resting-State Functional MRI

for UNC/UMN Baby Connectome Project Consortium

研究成果: 書籍/レポート タイプへの寄稿会議への寄与

17 被引用数 (Scopus)

抄録

Resting-state functional magnetic resonance imaging (rs-fMRI) is a non-invasive functional imaging technique that has been widely used to investigate brain functional connectome. Noises and artifacts are dominant in the raw rs-fMRI, making effective noise removal a necessity prior to any subsequent analysis. Without requiring any additional biophysiological recording devices, directly applying independent component analysis on rs-fMRI data becomes a popular process further separating structured noise from signals. However, fast and accurate automatic identification of the noise-related components is critical. Conventional machine learning techniques have been used in training such a classifier with manually engineered features of the components, which usually takes a long time even in the testing phase because its success relies on exhaustively extraction of spatial and temporal features and assembling multiple complicated classifiers to reach satisfactory results. In this paper, we proposed a novel, end-to-end, deep learning-based framework dedicated for noise component identification via effective, automatic, multilayer, hierarchically embedded feature extraction. The merit that does not require any assumptions on the features guarantees its unprecedented performance on the rs-fMRI data even from very heterogeneous cohorts. The speed of this method can be further accelerated due to its inherent ability of parallel computing with GPU. We validate our method with a challenging infant rs-fMRI dataset with high resolution and high quality, which are very different from the commonly used adult data. Our proposed method is more general, hypothesis-free, fast (<1 s for single component classification), and accurate (>97% accuracy).

本文言語英語
ホスト出版物のタイトルMedical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings
編集者Dinggang Shen, Pew-Thian Yap, Tianming Liu, Terry M. Peters, Ali Khan, Lawrence H. Staib, Caroline Essert, Sean Zhou
出版社Springer Science and Business Media Deutschland GmbH
ページ754-762
ページ数9
ISBN(印刷版)9783030322472
DOI
出版ステータス出版済み - 2019
外部発表はい
イベント22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 - Shenzhen, 中国
継続期間: 10月 13 201910月 17 2019

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
11766 LNCS
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

会議

会議22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019
国/地域中国
CityShenzhen
Period10/13/1910/17/19

!!!All Science Journal Classification (ASJC) codes

  • 理論的コンピュータサイエンス
  • コンピュータサイエンス一般

フィンガープリント

「A Deep Learning Framework for Noise Component Detection from Resting-State Functional MRI」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル