Fully automatic brain segmentation for Alzheimer's disease detection from magnetic resonance images

This paper proposes a new automatic method to segment the whole brain in magnetic resonance (MR) image series and calculate its volume for detecting Alzheimer's disease (AD). The underlying MR images were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. The wh...

Full description

Saved in:
Bibliographic Details
Main Authors: Chaturaphat Tanchi, Nipon Theera-Umpon, Sansanee Auephanwiriyakul
Format: Conference Proceeding
Published: 2018
Subjects:
Online Access:https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84877829460&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/51500
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
Description
Summary:This paper proposes a new automatic method to segment the whole brain in magnetic resonance (MR) image series and calculate its volume for detecting Alzheimer's disease (AD). The underlying MR images were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. The whole brain T1-weighted MRI was performed at 1.5 T in 100 subjects. The proposed automatic segmentation method is based on the mathematical morphology of image and our proposed technique called the 'brain template' to limit the boundary around the brain. The results show that the volumes of AD patients, mild cognitive impairment (MCI) patients, and normal persons are 828±49mm3, 922±30 mm3, and 1056±102 mm3, respectively. We also performed the three-class classification problem on the data set using the Bayes classifier and four-fold cross validation. The classification rate of 87% is achieved on the test sets. © 2012 IEEE.