Automatic segmentation of the hippocampus and the amygdala driven by hybrid constraints: method and validation.

The segmentation from MRI of macroscopically ill-defined and highly variable structures, such as the hippocampus (Hc) and the amygdala (Am), requires the use of specific constraints. Here, we describe and evaluate a fast fully automatic hybrid segmentation that uses knowledge derived from probabilistic atlases and anatomical landmarks, adapted from a semi-automatic method. The algorithm was designed at the outset for application on images from healthy subjects and patients with hippocampal sclerosis. Probabilistic atlases were built from 16 healthy subjects, registered using SPM5. Local mismatch in the atlas registration step was automatically detected and corrected. Quantitative evaluation with respect to manual segmentations was performed on the 16 young subjects, with a leave-one-out strategy, a mixed cohort of 8 controls and 15 patients with epilepsy with variable degrees of hippocampal sclerosis, and 8 healthy subjects acquired on a 3 T scanner. Seven performance indices were computed, among which error on volumes RV and Dice overlap K. The method proved to be fast, robust and accurate. For Hc, results with the new method were: 16 young subjects {RV=5%, K=87%}; mixed cohort {RV=8%, K=84%}; 3 T cohort {RV=9%, K=85%}. Results were better than with atlas-based (thresholded probability map) or semi-automatic segmentations. Atlas mismatch detection and correction proved efficient for the most sclerotic Hc. For Am, results were: 16 young controls {RV=7%, K=85%}; mixed cohort {RV=19%, K=78%}; 3 T cohort {RV=10%, K=77%}. Results were better than with the semi-automatic segmentation, and were also better than atlas-based segmentations for the 16 young subjects.

A longitudinal study of brain morphometrics using quantitative magnetic resonance imaging and difference image analysis.

Serial quantitative magnetic resonance imaging (MRI) allows the detection of subtle volumetric changes in brain volume. We used serial volumetry and voxel-based difference image analysis to quantify and characterize longitudinal changes in the hippocampus, cerebellum, and neocortex in younger and middle-age individuals. Paired volumetric MRI brain scans 3.5 years apart were performed on 90 healthy subjects 14 to 77 years old. Quantitative assessment of registered images included hippocampal volumetry, cerebellar volumetry, and automatically determined regional brain volumes. Longitudinal volume changes in three age epochs (<35, 35-54, >54 years) were compared and neocortical changes beyond regions of interest were visualized using filtered difference images. Cross-sectional analysis revealed a significant association between age and reduction in all brain volumes except hippocampal volume. Changes in normalized hippocampal and white matter volume were significantly different among the three groups. Individual analysis revealed 5 subjects with significant longitudinal volume changes lying outside the normative range. Difference image analysis showed global involutional changes in the >54 age group. Our findings suggest that cross-sectional observations in intracranial volume, cerebellar volume, and gray matter volume are likely to reflect uniform rates of volume loss or secular changes. Accelerated brain atrophy was seen from the age of 35-54 and increased rates of hippocampal atrophy from the age of 54. Our findings emphasize the importance of controlling for age effects when studying pathological brain changes over a wide age range.

Psychiatric adverse events during levetiracetam therapy.

The prevalence and psychopathologic features of psychiatric adverse events (PAE) in 517 patients taking levetiracetam (LEV) were investigated. Fifty-three (10.1%) patients developed PAE. A significant association was found with previous psychiatric history, history of febrile convulsions, and history of status epilepticus, whereas lamotrigine co-therapy had a protective effect. PAE were not related to the titration schedule of LEV, and certain patients seem to be biologically more vulnerable.