Show simple item record

dc.contributor.authorMeng, Meng
dc.date.accessioned2014-03-04T20:24:39Z
dc.date.available2014-03-04T20:24:39Z
dc.date.issued2011-12
dc.identifier.othermeng_meng_201112_ms
dc.identifier.urihttp://purl.galileo.usg.edu/uga_etd/meng_meng_201112_ms
dc.identifier.urihttp://hdl.handle.net/10724/27767
dc.description.abstractAlzheimer's is an irreversible brain disease that impairs memory, thinking and behavior and leads ultimately to death.Research has shown that individuals with MCI (mild cognitive impairment), the pre-stage of Alzheimer's, have an increased risk of developing Alzheimer's over the next few years . It is useful and important to diagnose and predict MCI's conversion to Alzheimer's as early as possible for appropriate treatment. In our study, we use numerous machine learning, feature selection as well as clustering methods for this prediction purpose. High precision of prediction is observed for both 10-fold and 2-fold cross-validation. We also use L1 and L2-norm shrinkage terms to control the model complexity. As a result, the prediction error is reduced. These findings illustrate that machine learning methods accurately and reliably predict MCI's conversion, and potentially provide a great assistance to medical diagnosis.
dc.languageeng
dc.publisheruga
dc.rightspublic
dc.subjectAlzheimer's disease
dc.subjectMRI
dc.subjectMild cognitive impairment
dc.subjectSVM
dc.subjectLogistic
dc.subjectLasso
dc.subjectLoss function
dc.subjectRegularization
dc.titleAutomated MRI prediction of Alzheimer's disease development by machine learning methods
dc.typeThesis
dc.description.degreeMS
dc.description.departmentComputer Science
dc.description.majorComputer Science
dc.description.advisorKhaled Rasheed
dc.description.committeeKhaled Rasheed
dc.description.committeeTianming Liu
dc.description.committeeHamid Arabnia


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record