Ralf Schönmeyer

The corpus callosum in schizophrenia-volume and connectivity changes affect specific regions

The corpus callosum (CC) is of great interest for pathophysiological models of schizophrenia. Volume and structural integrity of the CC have been examined by volumetric and diffusion tensor imaging (DTI) studies, but results were not consistent across methods or studies. A possible explanation may be varying methodologies and accuracy of measurements based on a single slice or small regions of interest. In addition, none of the studies examined volume and diffusion values in the same group of patients, and thus the relationship between these anatomical measures is not clear. We used an automatic algorithm to segment seven midline slices of the CC from DTI images. We compared volume and the DTI measures fractional anisotropy (FA) and mean diffusivity (MD) in the CC and its subdivisions in the schizophrenia patients and matched controls. Patients had decreased volume, decreased FA and increased MD of the whole CC. The important novel finding is, however, that not all regions were equally affected by anatomical changes. The results emphasize the importance of using different methods in evaluation of white matter (WM) in schizophrenia to avoid false negative findings. In addition, the measures were highly correlated with each other, implying a common pathological process influencing FA, MD and volume of the CC. Although we cannot rule out other mechanisms affecting volume, FA and MD, converging evidence from cytoarchitectonic and genetic studies suggests that WM changes observed in schizophrenia may involve disintegration of healthy, functional axons and strengthening of aberrant connections resulting in increased severity of clinical symptoms.

Reduced laterality as a trait marker of schizophrenia--evidence from structural and functional neuroimaging.

Laterality is a characteristic principle of the organization of the brain systems for language, and reduced hemispheric asymmetry has been considered a risk factor for schizophrenia. Here we sought support for the risk factor hypothesis by investigating whether reduced asymmetry of temporal lobe structure and function is also present in unaffected relatives. Sixteen schizophrenia patients, 16 age-matched first-degree relatives, and 15 healthy controls underwent high-resolution three-dimensional anatomical imaging and functional magnetic resonance imaging during auditory stimulation. Both the overall auditory cortex and planum temporale volumes and the lateralization to the left hemisphere were markedly reduced in patients. The decrease of lateralization correlated with increased severity of symptoms. In addition, both the overall functional activation in response to auditory stimulation and its asymmetry were reduced in the patients. Relatives had intermediate values between patients and controls on both structural and functional measures. This study provides added support for the idea that reduced hemispheric asymmetry is a biological risk factor for schizophrenia.

Interhemispheric hypoconnectivity in schizophrenia: Fiber integrity and volume differences of the corpus callosum in patients and unaffected relatives

Changes in hemispheric asymmetry and inter-hemispheric connectivity have been reported in schizophrenia. However, the genetic contribution to these alterations is still unclear. In the current study, we applied an automatic segmentation method to structural MRI and diffusion tensor imaging (DTI) data and examined volume and fiber integrity of the corpus callosum (CC), the main interhemispheric fiber tract, in 16 chronic schizophrenia (SZ) patients, matched first degree relatives and controls. SZ patients and relatives had smaller CC volumes than controls, particularly in the posterior genu, isthmus and splenium. Fractional anisotropy (FA), an indicator of fiber integrity, was reduced in patients and relatives in the whole CC, the inferior genu, the superior genu and the isthmus. Correspondingly, the mean diffusivity (MD) values of the whole CC and the isthmus were higher in patients and their unaffected relatives, indicating decreased compactness and increased intercellular space. Relatives had intermediate values in the volumetric and fiber integrity measurements between patients and controls. Lower CC volume and fiber integrity in SZ patients were associated with more severe auditory hallucinations. These results support the connectivity hypothesis of SZ (Friston, 1998) and particularly highlight the altered interhemispheric connectivity, which appears to be a genetic feature of SZ risk.

Mild Cognitive Impairment in the Elderly is Associated with Volume Loss of the Cholinergic Basal Forebrain Region

BACKGROUND Cholinergic neurons within the basal forebrain are assumed to be an early (preclinical) manifestation site of pathological changes in Alzheimers disease (AD). METHODS We used morphometric magnetic resonance imaging (MRI) to detect and quantify atrophic changes in the basal forebrain of subjects suffering from amnestic mild cognitive impairment (aMCI). Three Tesla magnetic resonance (MR) data of 26 aMCI patients, 46 cognitively normal elderly control subjects (CO), and 12 patients suffering from Alzheimers dementia were analyzed, including segmentation and quantification of brain tissue as well as a segmentation of basal forebrain structures (substantia innominata [SI]). RESULTS We found the volume of the SI to be significantly different between groups in that control subjects showed the largest SI volumes, followed by aMCI and AD patients. CONCLUSIONS These results are in line with the hypothesis that cell loss within the cholinergic basal forebrain regions occurs already in the early (predementia) stage of AD. In vivo quantification of these changes might be of use as a novel neuroimaging marker of cholinergic neurodegeneration in AD.

Multi-template training for image processing with cellular neural networks

Cellular neural networks (CNN) are often considered as massive parallel computing arrays for high speed image processing. In order to find appropriate CNN templates, optimization methods are necessary in many cases. We consider the optimization method Iterative Annealing directly using the output of a hardware realization of a CNN-UM Chip. The procedure presented in this contribution generates highly adapted sets of templates for complex image processing tasks. With this approach it is also possible to tune existing CNN programs to compensate inaccuracies of analog CNN hardware leading to noise reduction and more robust behaviour. Finally, an application of practical interest has been developed, by using the introduced method. We achieved the tracing of a certain selected object out of an image sequence showing many moving objects.

Automatisierte Segmentierung der Seitenventrikel des menschlichen Gehirns aus kernspintomographischen Datensätzen

Volumetrische Mase des menschlichen Ventrikelsystems sind ein wichtiger Bestandteil zur Beurteilung des Krankheitsverlaufs bei neurodegenerativen Krankheiten, wie z.B der Alzheimerschen Demenz (AD). In dieser Arbeit wird ein Segmentierungs-Algorithmus vorgestellt, der die Lage und das Volumen der Seitenventrikel des menschlichen Gehirns automatisch bestimmen kann, so das Zeit- und Arbeitsaufwand manueller Segmentierungen stark reduziert werden. Die verwendeten Algorithmen verfolgen dabei einen regelbasierten und objektorientieren Ansatz, der hier erstmals fur die Verwendung von 3D-Daten angepast und evaluiert wurde. Die zugrundeliegenden 20 Datensatze entstammen einer AD-Studie, wobei 10 Datensatze von 5 Experten zum Vergleich manuell segmentiert wurden. Die erzielten Ergebnisse liegen im Bereich der Inter-Rater-Variabilitat, und der voll-automatische Algorithmus scheitert nur in 1/20 der Falle.

Automatische Segmentierung des Corpus Callosum aus sagittalen Schichten von kernspintomographischen Datensätzen

In dieser Arbeit stellen wir einen voll-automatisierten Algorithmus vor, der in der Lage ist, die Struktur des Corpus Callosum aus sagittalen Schichten von T1-gewichteten kernspintomographischen Datensatzen des menschlichen Gehirns zu segmentieren. Die Segmentierungsergebnisse werden dabei fur die Untersuchung morphometrischer Merkmale in einer Studie zur Schizophrenie in definierte Abschnitte unterteilt, um sie im weiteren Verlauf statistisch auswerten zu konnen. Der Algorithmus wurde unter Zuhilfenahme der Cognition Network Technologie implementiert, die eine regelbasierte und kontextsensitive Handhabung der Bilddaten erlaubt, und dabei nur wenige Voraussetzungen uber die Beschaffenheit und Qualitat der zu verarbeitenden Datens atze macht. Das Verfahren scheitert im Rahmen einer Testreihe bei einem von 50 Datensatzen und erzielt ansonsten einen Dice-Koeffizienten von 0,97 im Vergleich zu manuell segmentierten Ergebnissen.

On-chip template training for pattern matching by cellular neural network universal machines (CNN-UM)

Pattern matching problems using statistical methods generally result in high computational effort. On the other side algorithms based on CNN technology can provide efficient new solutions for complex image processing tasks. In various applications template values are determined by an optimization procedure using simulation systems. In this contribution an optimization method directly interacting with a CNN-UM chip will be presented to treat a CNN based pattern matching problem. Thereby a certain binary pattern of an image also comprising other different patterns should be extracted. The proposed on-chip training leads to highly adapted templates solving the given tasks in different setups.

Automated Co-Analysis of MALDI and H&E Images of Retinal Tissue for an Improved Spatial MALDI Resolution

MALDI imaging is a powerful technology to gain proteome information with high mass spectroscopic resolution from tissue slides. As its spatial resolution is lower than that of standard optical microscopy, improving the resolution to allow investigation of cell-sized objects is highly desirable. In this contribution we present an approach to virtually improve MALDI’s spatial resolution in cases where the relevant structures have an approximately linear shape and can be transformed into a one-dimensional problem. By applying an automated image analysis to co-registered microscopy data, we can obtain the parameters necessary to support a MALDI-based modeling approach for investigating porcine retinal tissue.

Vollautomatische Segmentierung der weißen Hirnsubstanz oberhalb der Seitenventrikel aus kernspintomographischen Datensätzen

In dieser Arbeit stellen wir einen voll-automatisierten Algorithmus vor, der in der Lage ist, die weise Substanz in der Region oberhalb der Ventrikel aus T1-gewichteten anatomischen MR-Datensatzen des menschlichen Gehirns zu segmentieren. Im Gegensatz zu anderen Methoden werden dabei nur wenige Voraussetzungen uber die zu verarbeitenden Daten gemacht, und es sind keine ggf. manuell zu begleitenden Vorverarbeitungsschritte, wie z.B. Filterung und/oder Normalisierung, notig. Der Algorithmus wurde unter Zuhilfenahme der Cognition Network Technology implementiert, die eine wissensbasierte und kontextsensitive Handhabung der Bilddaten erlaubt. Eine quantitative Auswertung anhand von bis jetzt 10 Datensatzen zeigt, das im Vergleich zu einer manuellen Segmentierung eine overlay-metric von 0.9 erzielt wird.