Automated image analysis of disturbed cytoarchitecture in Brodmann area 10 in schizophrenia

• Published in: Schizophrenia research Vol. 62; no. 1; pp. 133 - 140
• Main Authors: Vogeley, Kai, Tepest, Ralf, Schneider-Axmann, Thomas, Hütte, Helge, Zilles, Karl, Honer, William G., Falkai, Peter
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.07.2003; Elsevier Science
• Subjects: Adult; Adult and adolescent clinical studies; Age Factors; Aged; Automated image analysis; Biological and medical sciences; Brodmann area 10; Cell Count; Cerebral Cortex - pathology; Cytoarchitectonic profiles; Dominance, Cerebral - physiology; Female; Gray-level index (GLI); Humans; Image Processing, Computer-Assisted; Male; Medical sciences; Microscopy; Middle Aged; Neurons - pathology; Neuropil - pathology; Prefrontal Cortex - pathology; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Psychoses; Reference Values; Schizophrenia; Schizophrenia - pathology; Schizophrenia - physiopathology; Sex Factors; Schizophrenia; Cytoarchitectonic profiles; Brodmann area 10; Gray-level index (GLI); Automated image analysis; Human; Methodology; Grey matter; Postmortem; Central nervous system; Prefrontal cortex; Automatic processing; Psychosis; Image analysis; Cell population; Microscopic structure; Quantitative analysis; Brain (vertebrata)
• ISSN: 0920-9964; 1573-2509
• DOI: 10.1016/S0920-9964(02)00325-0

To detect cytoarchitectonic abnormalities in the Brodmann area 10 (BA10) of schizophrenic patients, we applied a newly modified variant of the gray-level index (GLI) method as fully automated image analysis method providing cytoarchitectonic profiles of the whole cortex as a scanning tool. Microscopic images of silver-stained sections of 20 schizophrenic brains compared to 20 control brains were automatically scanned and binarized at an adaptive threshold. In 30 measuring fields through the whole cortical depth, the dependent measure of gray-level index (GLI) as the area-percentage covered by perikarya in a measuring field was obtained providing a cytoarchitectonic profile. GLI is an estimate of the volume density of perikarya. A statistical analysis of mean GLI values was performed for six compartments, separately, approximately corresponding to cortical layers. Results revealed significant GLI reductions in schizophrenic brains in all six compartments suggesting either a decreased perikarya fraction or an increased neuropil fraction. The described automated image analysis method providing cytoarchitectonic profiles can be applied as a fast and observer-independent scanning tool to detect cytoarchitectonic abnormalities in multiple brain regions.