Technical Challenges in the Statistical Analysis of Neuroimaging Studies of Alzheimer's Disease
The neuroimaging field faces significant challenges in the study of neurodegenerative diseases such as Alzheimer's Disease (AD): it utilizes various imaging techniques and software packages that were originally developed and characterized for neuroimaging studies of healthy brains. These software often require extensive in-house modification that make analysis pipelines opaque and very hard to reproduce even in healthy populations. Additionally, automated segmentations of the brain are very difficult in the presence of brain pathologies such as white matter hyperintensities (WMH). WMHs appear as hyperintense areas in magnetic resonance imaging (MRI) and are frequently found in AD population's brain. WMHs present a challenge for standard segmentation algorithms that misclassify WMHs as gray matter (GM) and for co-registration with other imaging modalities such as Positron Emission Tomography (PET) that is used to detect amyloid plague in the AD brains. An overview of reproducibility in the context of healthy brain is presented along with some newly developed improvements for segmentation methods in presence of WMHs in AD. A high dimensional longitudinal data analysis model is introduced. The proposed methods showed significant improvement in tissue classification for AD.
Eloyan A et al. (2014). Health Effects of Lesion Localization in Multiple Sclerosis:Spatial Registration and Confounding Adjustment. PLoS ONE, doi:10.1371/journal.pone.0107263.
Karim HT et al. (2016). The effects of white matter disease on the accuracy of automated segmentation. Psychiatry Res, 14, 253-257.
Tudorascu DL et al. (2016). Reproducibility and Bias in Healthy Brain Segmentation: Comparison of Two Popular Neuroimaging Platforms. Frontiers in Neuroscience, 10, 503-510.
Wu M et al. (2006). A fully automated method for quantifying and localizing white matter hyperintensities on MR images. Psychiatry Res, 144, 133-142.