TY - GEN
T1 - A study of data fusion for Alzheimer's disease based on diffusion magnetic resonance imaging
AU - Zhang, Changle
AU - Mao, Shuai
AU - Wong, Chunsing
AU - Hui, Edward S.
AU - Ye, Chenfei
AU - Li, Hengtong
AU - Ma, Jingbo
AU - Ma, Heather T.
N1 - Publisher Copyright:
© 2016 IEEE.
Copyright:
Copyright 2017 Elsevier B.V., All rights reserved.
PY - 2017/2/8
Y1 - 2017/2/8
N2 - Alzheimer's disease (AD) has become one of the most serious healthcare problems. As a result, early diagnosis for AD is important for intervention of its deterioration at an early stage. Diffusion tensor imaging (DTI) provides a non-intrusive examination of cranial nerve diseases, but the precise quantification is a problem for diagnosis. In current study, we proposed an AD recognition method based on quantitative analysis and data fusion of T1 and DTI images. Through segmentation on T1 image, feature extraction of DTI data, and data fusion based on multi-features, recognition accuracy of AD can attain as high as 95.4%. For details, patterns of different features from DTI data were investigated showing that the fusion of volume values, DTI and diffusion kurtosis imaging (DKI) parameters and the Gaussian Mixture Model (GMM) parameters of DTI and DKI are the most distinctive individual features in AD patients. It also implies that during the AD development, neural degeneration could be demyelination, reduction of structural complexity and decrease of brain microstructure connectivity, which can be reflected by different quantitative parameters from brain images, such as brain structure volume, DTI parameters and DKI parameters. Further study on the image features, including more image types, would provide valued method for AD screening or early diagnosis.
AB - Alzheimer's disease (AD) has become one of the most serious healthcare problems. As a result, early diagnosis for AD is important for intervention of its deterioration at an early stage. Diffusion tensor imaging (DTI) provides a non-intrusive examination of cranial nerve diseases, but the precise quantification is a problem for diagnosis. In current study, we proposed an AD recognition method based on quantitative analysis and data fusion of T1 and DTI images. Through segmentation on T1 image, feature extraction of DTI data, and data fusion based on multi-features, recognition accuracy of AD can attain as high as 95.4%. For details, patterns of different features from DTI data were investigated showing that the fusion of volume values, DTI and diffusion kurtosis imaging (DKI) parameters and the Gaussian Mixture Model (GMM) parameters of DTI and DKI are the most distinctive individual features in AD patients. It also implies that during the AD development, neural degeneration could be demyelination, reduction of structural complexity and decrease of brain microstructure connectivity, which can be reflected by different quantitative parameters from brain images, such as brain structure volume, DTI parameters and DKI parameters. Further study on the image features, including more image types, would provide valued method for AD screening or early diagnosis.
UR - http://www.scopus.com/inward/record.url?scp=85015408927&partnerID=8YFLogxK
U2 - 10.1109/TENCON.2016.7848160
DO - 10.1109/TENCON.2016.7848160
M3 - Conference article published in proceeding or book
AN - SCOPUS:85015408927
T3 - IEEE Region 10 Annual International Conference, Proceedings/TENCON
SP - 1019
EP - 1022
BT - Proceedings of the 2016 IEEE Region 10 Conference, TENCON 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2016 IEEE Region 10 Conference, TENCON 2016
Y2 - 22 November 2016 through 25 November 2016
ER -