Spatially augmented LPboosting for AD classification with evaluations on the ADNI dataset

Abstract

Structural and functional brain images are playing an important role in helping us understand the changes associated with neurological disorders such as Alzheimer's disease (AD). Recent efforts have now started investigating their utility for diagnosis purposes. This line of research has shown promising results where methods from machine learning (such as Support Vector Machines) have been used to identify AD-related patterns from images, for use in diagnosing new individual subjects. In this paper, we propose a new framework for AD classification which makes use of the Linear Program (LP) boosting with novel additional regularization based on spatial "smoothness" in 3D image coordinate spaces. The algorithm formalizes the expectation that since the examples for training the classifier are images, the voxels eventually selected for specifying the decision boundary must constitute spatially contiguous chunks, i.e., "regions" must be preferred over isolated voxels. This prior belief turns out to be useful for significantly reducing the space of possible classifiers and leads to substantial benefits in generalization. In our method, the requirement of spatial contiguity (of selected discriminating voxels) is incorporated within the optimization framework directly. Other methods have made use of similar biases as a pre-or post-processing step, however, our model incorporates this emphasis on spatial smoothness directly into the learning step. We report on extensive evaluations of our algorithm on MR and FDG-PET images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, and discuss the relationship of the classification output with the clinical and cognitive biomarker data available within ADNI.

Combining MRI and CSF measures for classification of Alzheimer's disease and prediction of mild cognitive impairment conversion

“…Previous studies have combined different biomarkers for classification of individual subjects using MRI and FDG-PET (Fan et al, 2008;Hinrichs et al, 2009aHinrichs et al, , 2009b, MRI and CSF (Davatzikos et al, 2011;Kohannim et al, 2010;Nettiksimmons et al, 2010), as well as MRI, FDG-PET and CSF (Kohannim et al, 2010;Zhang et al, 2011). In the current manuscript we used the multivariate OPLS method with MRI and CSF data as input.…”

Automatic classification of patients with Alzheimer's disease from structural MRI: A comparison of ten methods using the ADNI database

“…Recently, several approaches have been proposed to automatically classify patients with AD and/or MCI from anatomical MRI (Fan et al, 2005(Fan et al, , 2007Colliot et al, 2008;Davatzikos et al, 2008a,b;Klöppel et al, 2008;Vemuri et al, 2008;Chupin et al, 2009a,b;Desikan et al, 2009;Gerardin et al, 2009;Hinrichs et al, 2009;Magnin et al, 2009;Misra et al, 2009;Querbes et al, 2009). These approaches could have the potential to assist in the early diagnosis of AD.…”

“…In the first category, the features are defined at the level of the MRI voxel. Specifically, the features are the probability of the different tissue classes (grey matter, white matter and cerebrospinal fluid) in a given voxel (Lao et al, 2004;Fan et al, 2007Fan et al, , 2008aDavatzikos et al, 2008a,b;Klöppel et al, 2008;Vemuri et al, 2008;Hinrichs et al, 2009;Magnin et al, 2009;Misra et al, 2009). Klöppel et al (2008) directly classified these features with an SVM.…”

Identifying Informative Imaging Biomarkers via Tree Structured Sparse Learning for AD Diagnosis

“…Neuroimaging measurements, including magnetic resonance image (MRI) and fluorodeoxyglucose positron emission tomography (FDG-PET), provide a powerful in vivo tool for helping diagnosis and longitudinal follow-up study of AD and MCI (Desikan et al 2009; Klöppel et al 2008; Stonnington et al 2010; Zhou et al 2011; Oliveira et al 2010; Leung et al 2010; Davatzikos et al 2010; Querbes et al 2009; Filipovych and Davatzikos 2011; Duchesne et al 2009; Fan et al 2007a; Wee et al 2011, 2012; Zhang and Shen et al 2012a, b; Li et al 2012). In recent years, a substantial research effort has been made to investigate many machine learning and pattern recognition technologies, such as sparse learning and support vector machines (SVM), in neuroimaging analysis to assist AD diagnosis (Davatzikos et al 2008a; Magnin et al 2009; Hinrichs et al 2009; Cuingnet et al 2011; Wolz et al 2011; Liu et al 2012a, b). Various methods have been proposed for processing and analysis of neuroimages to investigate the pathological changes related to brain diseases (Xue et al 2006; Wu et al 2006; Yang et al 2008; Magnin et al 2009; Shen et al 1999; Jia et al 2010, Tang et al 2009).…”

“…The existence of structural relationship in spatial features can be used to build better feature selection method (Hinrichs et al 2009). Recently, group Lasso was extended from the L1-norm Lasso to find solutions that are sparse on the group level of features (Yuan and Lin 2006).…”