Alzheimer's disease diagnosis in individual subjects using structural MR images: Validation studies

• Published in: NeuroImage (Orlando, Fla.) Vol. 39; no. 3; pp. 1186 - 1197
• Main Authors: Vemuri, Prashanthi, Gunter, Jeffrey L., Senjem, Matthew L., Whitwell, Jennifer L., Kantarci, Kejal, Knopman, David S., Boeve, Bradley F., Petersen, Ronald C., Jack, Clifford R.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.02.2008; Elsevier Limited
• Subjects: Age; Aged; Aged, 80 and over; Algorithms; Alleles; Alzheimer Disease - classification; Alzheimer Disease - diagnosis; Alzheimer Disease - genetics; Alzheimer's; Alzheimer's disease; Apolipoproteins E - genetics; Artificial Intelligence; Cerebral Ventricles - pathology; Classification; Demographics; Diagnosis; Female; Gender; Hippocampus - pathology; Humans; Magnetic Resonance Imaging; Male; Middle Aged; Models, Statistical; Multivariate analysis; NMR; Nonlinear Dynamics; Nuclear magnetic resonance; Patients; Reproducibility of Results; Studies; Support vector machines; Support vector machines; Diagnosis; Classification; Alzheimer's
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2007.09.073

To develop and validate a tool for Alzheimer's disease (AD) diagnosis in individual subjects using support vector machine (SVM)-based classification of structural MR (sMR) images. Libraries of sMR scans of clinically well characterized subjects can be harnessed for the purpose of diagnosing new incoming subjects. One hundred ninety patients with probable AD were age- and gender-matched with 190 cognitively normal (CN) subjects. Three different classification models were implemented: Model I uses tissue densities obtained from sMR scans to give STructural Abnormality iNDex (STAND)-score; and Models II and III use tissue densities as well as covariates (demographics and Apolipoprotein E genotype) to give adjusted-STAND (aSTAND)-score. Data from 140 AD and 140 CN were used for training. The SVM parameter optimization and training were done by four-fold cross validation (CV). The remaining independent sample of 50 AD and 50 CN was used to obtain a minimally biased estimate of the generalization error of the algorithm. The CV accuracy of Model II and Model III aSTAND-scores was 88.5% and 89.3%, respectively, and the developed models generalized well on the independent test data sets. Anatomic patterns best differentiating the groups were consistent with the known distribution of neurofibrillary AD pathology. This paper presents preliminary evidence that application of SVM-based classification of an individual sMR scan relative to a library of scans can provide useful information in individual subjects for diagnosis of AD. Including demographic and genetic information in the classification algorithm slightly improves diagnostic accuracy.