Title: Probabilistic Framework for Population Analysis of Brain MR Images (Probabilistische raamwerk voor populatie analyse van MR hersenbeelden)
Other Titles: Probabilistic Framework for Population Analysis of Brain MR Images
Authors: Ribbens, Annemie
Issue Date: 13-Dec-2012
Abstract: Many neurodegenerative diseases such as Alzheimer¬ís disease can be chara cterized by their gradual modification of the cellular environment resul ting in macroscopic brain changes. Groupwise analysis of large sets of m agnetic resonance (MR) images, visualizing the morphology of the human b rain, can provide reliable measurable image features indicative for a sp ecific disease and/or disease stage. Such features, called biomarkers, c an contribute to early diagnosis, to the assessment of therapy response as well as to the study of the underlying causes of the neurodegenerativ e disease. In this PhD, we developed a Bayesian framework for both cross-sectional and longitudinal groupwise analysis of a collection of brain MR images. Contrary to most state-of-the-art strategies for cross-sectional analysi s, which compare homogeneous subgroups of images selected based on prior clinical knowledge, our framework can handle heterogeneous collections of images, for instance of both normal controls as well as patients in d ifferent disease states. The proposed framework simultaneously performs segmentation of each image into the major tissue classes (white matter, gray matter, CSF) and automatically clusters all images of the set in ho mogeneous subgroups based on the image morphology extracted from the ima ge segmentations. A probabilistic brain atlas is iteratively constructed for each cluster within the framework. These atlases visualize the mean morphology of the subgroup and support the clustering and segmentation processes. The constructed morphological subgroups can be correlated wit h the clinical diagnosis of each subject, while the clustering process r eveals the distinctive cluster-specific image features globally and for each individual image. The framework is adapted to take additional clini cal knowledge into account in the clustering process if desired. The lon gitudinal framework is a direct extension of the cross-sectional framewo rk, analyzing multi-temporal brain MR image sequences of different subje cts suffering from the same neurodegenerative disease, aiming to assess both the patient-specific and the population-wise disease evolution base d on the structural changes over time. The clustering process reveals th e subject-specific disease evolution, while the constructed atlases form a longitudinal atlas representing the temporal changes in brain morphol ogy caused by the disorder. A proof-of-concept of the framework is demonstrated on synthetic images and its feasibility is further shown on brain MR images from different p ublicly available data sets, including the simulated BrainWeb data set a s well as the large international initiatives ADNI and OASIS. The experi ments show that our framework provides accurate, more unbiased (and temp orally consistent) image segmentations and that it can detect the clinic ally relevant morphological subgroups (cross-sectionally) or the individ ual disease progression (longitudinally) in a set of images or image seq uences. The constructed atlases indicate the disease (stage) morphology better than traditional tools for cross-sectional and longitudinal analy sis based on clinical prior knowledge. Finally, it is illustrated that o ur combinational approach for segmentation and clustering can find disti nctive image features that would not have been picked up by handling bot h processes separately and subsequently. In conclusion, the presented method combines segmentation, clustering an d atlas construction in a unified probabilistic framework such that all techniques can cooperate and such that the method becomes more data-driv en. We provide initial evidence that our method facilitates a better rep resentation of the disease evolution, the detection of novel subgroups a nd the detection of novel disease-specific features. Therefore, our meth od can become an important tool for the development of novel and refined imaging biomarkers, providing new insights in structural change and dev elopment.
Description: Ribbens A., ''Probabilistic framework for population analysis of brain MR images'', Proefschrift voorgedragen tot het behalen van het doctoraat in de ingenieurswetenschappen, KU Leuven, December 2012, Leuven, Belgium.
Publication status: published
KU Leuven publication type: TH
Appears in Collections:ESAT - PSI, Processing Speech and Images

Files in This Item:
File Status SizeFormat
thesis.pdf Published 8209KbAdobe PDFView/Open

These files are only available to some KU Leuven staff members


All items in Lirias are protected by copyright, with all rights reserved.