Automated Detection of White Matter Changes in Elderly People Using Fuzzy, Geostatistical, and Information Combining Models
ABSTRACT Detection of white matter changes of the brain using magnetic resonance imaging (MRI) has increasingly been an active and challenging research area in computational neuroscience. There have rarely been any single image analysis methods that can effectively address the issue of automated quantification of neuroimages, which are subject to different interests of various medical hypotheses. This paper presents new image segmentation models for automated detection of white matter changes of the brain in an elderly population. The methods are based on the computational models of fuzzy clustering, possibilistic clustering, geostatistics, and knowledge combination. Experimental results on MRI data have shown that the proposed image analysis methodology can be applied as a very useful computerized tool for the validation of our particular medical question, where white matter changes of the brain are thought to be the most important social medical evidence.
- SourceAvailable from: Lena Bronge[show abstract] [hide abstract]
ABSTRACT: MRI is more sensitive than CT for detection of age-related white matter changes (ARWMC). Most rating scales estimate the degree and distribution of ARWMC either on CT or on MRI, and they differ in many aspects. This makes it difficult to compare CT and MRI studies. To be able to study the evolution and possible effect of drug treatment on ARWMC in large patient samples, it is necessary to have a rating scale constructed for both MRI and CT. We have developed and evaluated a new scale and studied ARWMC in a large number of patients examined with both MRI and CT. Seventy-seven patients with ARWMC on either CT or MRI were recruited and a complementary examination (MRI or CT) performed. The patients came from 4 centers in Europe, and the scans were rated by 4 raters on 1 occasion with the new ARWMC rating scale. The interrater reliability was evaluated by using kappa statistics. The degree and distribution of ARWMC in CT and MRI scans were compared in different brain areas. Interrater reliability was good for MRI (kappa=0.67) and moderate for CT (kappa=0.48). MRI was superior in detection of small ARWMC, whereas larger lesions were detected equally well with both CT and MRI. In the parieto-occipital and infratentorial areas, MRI detected significantly more ARWMC than did CT. In the frontal area and basal ganglia, no differences between modalities were found. When a fluid-attenuated inversion recovery sequence was used, MRI detected significantly more lesions than CT in frontal and parieto-occipital areas. No differences were found in basal ganglia and infratentorial areas. We present a new ARWMC scale applicable to both CT and MRI that has almost equal sensitivity, except for certain regions. The interrater reliability was slightly better for MRI, as was the detectability of small lesions.Stroke 07/2001; 32(6):1318-22. · 6.16 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Fuzzy clustering has been shown to be advantageous over crisp (or traditional) clustering methods in that total commitment of a vector to a given class is not required at each image pattern recognition iteration. Recently fuzzy clustering methods have shown spectacular ability to detect not only hypervolume clusters, but also clusters which are actually 'thin shells', i.e., curves and surfaces. Most analytic fuzzy clustering approaches are derived from the 'Fuzzy C-Means' (FCM) algorithm. The FCM uses the probabilistic constraint that the memberships of a data point across classes sum to one. This constraint was used to generate the membership update equations for an iterative algorithm. Recently, we cast the clustering problem into the framework of possibility theory using an approach in which the resulting partition of the data can be interpreted as a possibilistic partition, and the membership values may be interpreted as degrees of possibility of the points belonging to the classes. We show the ability of this approach to detect linear and quartic curves in the presence of considerable noise.06/1993;
- [show abstract] [hide abstract]
ABSTRACT: The impact of single and combined effects of subcortical white matter lesions (WMLs) and magnetic resonance imaging (MRI)-defined brain infarct on activities of daily living (ADL), depression, and health status perception was analyzed in community-dwelling elderly individuals. The study included 268 participants from the Memory and Morbidity in Augsburg Elderly (MEMO) project, a population-based study on individuals aged 65 to 83 years, conducted in Augsburg, Germany. Cerebral MRI was performed, and 2 geriatric performance tests, scales to assess ADL, depressive symptoms, and self-perceived health status were assessed. The prevalence of large (>10 mm) subcortical WML was 37.7% and of MRI-defined infarct-like lesions was 15.3%. Both vascular lesion types combined were found in 9% of the participants. Large WMLs were associated with significantly more impairments in basic ADL, inferior results in the performance tests, and a worse self-perceived health status compared to those without large WML. Magnetic resonance imaging-defined brain infarct was associated with impairments in performance tests. Participants with both lesion types were limited in all domains and were 2 to 3 times more likely to have impairments in all examined functions. Their risk of impairment in a specific function was considerably higher than the sum of the single risks associated with each lesion type alone. This study suggests that the single and especially the combined occurrence of common vascular brain lesions are associated with functional impairment. Identifying individuals with severe WML combined with MRI-defined brain infarct can help better understand the development of marked impairments in old age.Journal of Geriatric Psychiatry and Neurology 08/2009; 22(4):266-73. · 3.53 Impact Factor