Multi-region analysis of longitudinal FDG-PET for the classification of Alzheimer's disease

Abstract

Imaging biomarkers for Alzheimer’s disease are desirable for improved diagnosis and monitoring, as well as drug discovery. Automated image-based classification of individual patients could provide valuable diagnostic support for clinicians, when considered alongside cognitive assessment scores. We investigate the value of combining cross-sectional and longitudinal multi-region FDG-PET information for classification, using clinical and imaging data from the Alzheimer’s Disease Neuroimaging Initiative. Whole-brain segmentations into 83 anatomically defined regions were automatically generated for baseline and 12-month FDG-PET images. Regional signal intensities were extracted at each timepoint, as well as changes in signal intensity over the follow-up period. Features were provided to a support vector machine classifier. By combining 12-month signal intensities and changes over 12 months, we achieve significantly increased classification performance compared with using any of the three feature sets independently. Based on this combined feature set, we report classification accuracies of 88% between patients with Alzheimer’s disease and elderly healthy controls, and 65% between patients with stable mild cognitive impairment and those who subsequently progressed to Alzheimer’s disease. We demonstrate that information extracted from serial FDG-PET through regional analysis can be used to achieve state-of-the-art classification of diagnostic groups in a realistic multi-centre setting. This finding may be usefully applied in the diagnosis of Alzheimer’s disease, predicting disease course in individuals with mild cognitive impairment, and in the selection of participants for clinical trials.

Progressive Disintegration of Brain Networking from Normal Aging to Alzheimer Disease: Analysis of Independent Components of 18F-FDG PET Data

“…This observation confirms the utility of such data-driven methodology to uncover correlations that have pathophysiologic meaning (Table 3). Several studies have implemented automated image-based classification methods to differentiate AD and MCI patients from controls and have found such methods to have a statistical accuracy of 90% in discriminating AD patients from controls (25)(26)(27). When MCI patients who later converted to AD were investigated by unimodal biomarkers, the reported ability to discriminate from controls ranged from 80% (28) to 91% (29).…”

The Alzheimer's Disease Neuroimaging Initiative: A review of papers published since its inception

“…Mormino et al used 11 C-PiB PET imaging to deduce a cutoff point to optimally separate PiB-positive from PiB-negative MCI patients, and found that PiB-positive MCI patients had lower hippocampal volumes and greater episodic memory loss compared with MCI patients with 11 C-PiB levels below the cutoff point of 1.465. The addition of longitudinal data to baseline data to improve classification accuracy from anatomically selected features of FDG-PET scans was the approach taken by Gray et al [285]. Across all categories, improved classification accuracies were reported, ranging from 65% in the MCI converter versus non-converter classification to 88% in discriminating between control and AD patients (Table 8).…”

Application of Machine Learning to Arterial Spin Labeling in Mild Cognitive Impairment and Alzheimer Disease

“…Arterial spin labeling (ASL) MRI is a non-invasive, rapid and increasingly widely available method for quantifying CBF; ASL represents a potential alternative modality for measuring brain perfusion as compared to positron emission tomography (PET) [12][13][14] that may facilitate routine clinical application in the work-up of dementia. AD-associated perfusion changes measured by ASL are strongly correlated with glucose metabolism alterations as measured by PET [9,15,16].…”