A. Van der Linden

Watershed-based segmentation of 3D MR data for volume quantization

The aim of this work is the development of a semiautomatic segmentation technique for efficient and accurate volume quantization of Magnetic Resonance (MR) data. The proposed technique uses a 3D variant of Vincent and Soilles immersion-based watershed algorithm that is applied to the gradient magnitude of the MR data and that produces small volume primitives. The known drawback of the watershed algorithm, oversegmentation, is strongly reduced by a priori application of a 3D adaptive anisotropic diffusion filter to the MR data. Furthermore, oversegmentation is a posteriori reduced by properly merging small volume primitives that have similar gray level distributions. The outcome of the proceeding image processing steps is presented to the user for manual segmentation. Through selection of volume primitives, the user quickly segments of first slice, which contains the object of interest. Afterwards, the subsequent slices are automatically segmented by extrapolation. Segmentation results are contingently manually corrected. The proposed segmentation technique is tested on phantom objects, where segmentation errors less than 2% are observed. In addition, the technique is demonstrated on 3D MR data of the mouse head from which the cerebellum is extracted. Volumes of the mouse cerebellum and the mouse brains in toto are calculated.

Noninvasive in vivo MRI detection of neuritic plaques associated with iron in APP[V717I] transgenic mice, a model for Alzheimer's disease

Transgenic mice overexpressing the London mutant of human amyloid precursor protein (APP[V717I]) in neurons develop amyloid plaques in the brain, thus demonstrating the most prominent neuropathological hallmark of Alzheimers disease. In vivo 3D T2*‐weighted MRI on these mice (24 months of age) revealed hypointense brain inclusions that affected the thalamus almost exclusively. Upon correlating these MRI observations with a panel of different histologic staining techniques, it appeared that only plaques that were positive for both thioflavin‐S and iron were visible on the MR images. Numerous thioflavin‐S‐positive plaques in the cortex that did not display iron staining remained invisible to MRI. The in vivo detection of amyloid plaques in this mouse model, using the intrinsic MRI contrast arising from the iron associated with the plaques, creates an unexpected opportunity for the noninvasive investigation of the longitudinal development of the plaques in the same animal. Thus, this work provides further research opportunities for analyzing younger APP[V717I] mouse models with the knowledge of the final outcome at 24 months of age. Magn Reson Med 53:607–613, 2005.

Comparing BOLD fMRI signal changes in the awake and anesthetized rat during electrical forepaw stimulation

The difference between awake curarized and alpha-chloralose anesthetized animals was studied with respect to the BOLD signal response in an fMRI experiment. By studying the activation of the cortex upon electrical forepaw stimulation in the same rat, but following consecutively applied curarization and alpha-chloralose anesthesia protocols, it was possible to compare quantitatively the effect of both immobilization protocols on the fMRI data. The largest BOLD signal change as a result of forepaw stimulation was found in the awake condition, however the activated areas are less specific than those in the anesthetized state leaving it more difficult to interpret.

In vivo manganese-enhanced magnetic resonance imaging reveals connections and functional properties of the songbird vocal control system

Injection of manganese (Mn(2+)), a paramagnetic tract tracing agent and calcium analogue, into the high vocal center of starlings labeled within a few hours the nucleus robustus archistriatalis and area X as observed by in vivo magnetic resonance imaging. Structures highlighted by Mn(2+) accumulation assumed the expected tri-dimensional shape of the nucleus robustus archistriatalis and area X as identified by classical histological or neurochemical methods. The volume of these nuclei could be accurately calculated by segmentation of the areas highlighted by Mn(2+). Besides confirming previously established volumetric sex differences, Mn(2+) uptake into these nuclei revealed new functional sex differences affecting Mn(2+) transport. A faster transport was observed in males than in females and different relative amounts of Mn(2+) were transported to nucleus robustus archistriatalis and area X in males as compared to females. This new in vivo approach, allowing repeated measures, opens new vistas to study the remarkable seasonal plasticity in size and activity of song-control nuclei and correlate neuronal activity with behavior. It also provides new insights on in vivo axonal transport and neuronal activity in song-control nuclei of oscines.

Overexpression of human wildtype torsinA and human ΔGAG torsinA in a transgenic mouse model causes phenotypic abnormalities

Primary torsion dystonia is an autosomal-dominant inherited movement disorder. Most cases are caused by an in-frame deletion (GAG) of the DYT1 gene encoding torsinA. Reduced penetrance and phenotypic variability suggest that alteration of torsinA amino acid sequence is necessary but not sufficient for development of clinical symptoms and that additional factors must contribute to the factual manifestation of the disease. We generated 4 independent transgenic mouse lines, two overexpressing human mutant torsinA and two overexpressing human wildtype torsinA using a strong murine prion protein promoter. Our data provide for the first time in vivo evidence that not only mutant torsinA is detrimental to neuronal cells but that also wildtype torsinA can lead to neuronal dysfunction when overexpressed at high levels. This hypothesis is supported by (i) neuropathological findings, (ii) neurochemistry, (iii) behavioral abnormalities and (iv) DTI-MRI analysis.

Neonatal neuronal overexpression of glycogen synthase kinase-3β reduces brain size in transgenic mice

Glycogen synthase kinase-3beta (GSK-3beta) is important in neurogenesis. Here we demonstrate that the kinase influenced post-natal maturation and differentiation of neurons in vivo in transgenic mice that overexpress a constitutively active GSK-3beta[S9A]. Magnetic resonance imaging revealed a reduced volume of the entire brain, concordant with a nearly 20% reduction in wet brain weight. The reduced volume was most prominent for the cerebral cortex, without however, disturbing the normal cortical layering. The resulting compacted architecture was further demonstrated by an increased neuronal density, by reduced size of neuronal cell bodies and of the somatodendritic compartment of pyramidal neurons in the cortex. No evidence for apoptosis was obtained. The marked overall reduction in the level of the microtubule-associated protein 2 in brain and in spinal cord, did not affect the ultrastructure of the microtubular cytoskeleton in the proximal apical dendrites. The overall reduction in size of the entire CNS induced by constitutive active GSK-3beta caused only very subtle changes in the psychomotoric ability of adult and ageing GSK-3beta transgenic mice.

Restoration of MR-induced artifacts in simultaneously recorded MR/EEG data

During a Magnetic Resonance sequence, simultaneously acquired ElectroEncephaloGraphy (EEG) data are compromised by severe pollution due to artifacts originating from the switching of the magnetic field gradients. In this work, it is shown how these artifacts can be strongly reduced or even removed through application of an adaptive artifact restoration scheme. The method has proved to be fully automatic and to retain high frequency EEG information, which is indispensable for many EEG applications.

Mathematical framework for simulating diffusion tensor MR neural fiber bundles.

White matter (WM) fiber tractography (i.e., the reconstruction of the 3D architecture of WM fiber pathways) is known to be an important application of diffusion tensor magnetic resonance imaging (DT‐MRI). For the quantitative evaluation of several fiber‐tracking properties, such as accuracy, noise sensitivity, and robustness, synthetic ground‐truth DT‐MRI data are required. Moreover, an accurate simulated phantom is also required for optimization of the user‐defined tractography parameters, and objective comparisons between fiber‐tracking algorithms. Therefore, in this study a mathematical framework for simulating DT‐MRI data, based on the physical properties of WM fiber bundles, is presented. We obtained a model of a WM fiber bundle by parameterizing the various features that characterize this bundle. We then evaluated three different synthetic DT‐MRI models using experimental data in order to test the proposed methodology, and to determine the optimum model and parameter settings for constructing a realistic simulated DT‐MRI phantom. Several examples of how the mathematical framework can be applied to compare fiber‐tracking algorithms are presented. Magn Reson Med 53:944–953, 2005.

Reduction of ECG and gradient related artifacts in simultaneously recorded human EEG/MRI data.

Nowadays, electroencephalography signals can be acquired from a patient lying in a magnetic resonance imaging system. It is even possible to acquire EEG signals during an MR imaging sequence. However, such EEG signals are severely distorted by artifacts originating from various effects (e.g., MR gradients, ECG). In this paper, a simple method is presented to reduce such artifacts. Thereby, special attention is focused on artifacts related to the patients electrocardiogram. The method is shown to be effective, adaptive, and automatic.

Adaptive anisotropic noise filtering for magnitude MR data

Conventional noise filtering schemes applied to magnitude magnetic resonance (MR) images tacitly assume Gauss distributed noise. Magnitude MR data, however, are Rice distributed. Not incorporating this knowledge leads inevitably to biased results, in particular when applying such filters in regions with low signal-to-noise ratio. In this work, we show how the Rice data probability distribution can be incorporated so as to construct a noise filter that is far less biased.