Convolutional neural networks for cytoarchitectonic brain mapping at large scale.

Human brain atlases provide spatial reference systems for data characterizing brain organization at different levels, coming from different brains. Cytoarchitecture is a basic principle of the microstructural organization of the brain, as regional differences in the arrangement and composition of neuronal cells are indicators of changes in connectivity and function. Automated scanning procedures and observer-independent methods are prerequisites to reliably identify cytoarchitectonic areas, and to achieve reproducible models of brain segregation. Time becomes a key factor when moving from the analysis of single regions of interest towards high-throughput scanning of large series of whole-brain sections. Here we present a new workflow for mapping cytoarchitectonic areas in large series of cell-body stained histological sections of human postmortem brains. It is based on a Deep Convolutional Neural Network (CNN), which is trained on a pair of section images with annotations, with a large number of un-annotated sections in between. The model learns to create all missing annotations in between with high accuracy, and faster than our previous workflow based on observer-independent mapping. The new workflow does not require preceding 3D-reconstruction of sections, and is robust against histological artefacts. It processes large data sets with sizes in the order of multiple Terabytes efficiently. The workflow was integrated into a web interface, to allow access without expertise in deep learning and batch computing. Applying deep neural networks for cytoarchitectonic mapping opens new perspectives to enable high-resolution models of brain areas, introducing CNNs to identify borders of brain areas.

Cytoarchitectonic Maps of the Human Metathalamus in 3D Space.

The human metathalamus plays an important role in processing visual and auditory information. Understanding its layers and subdivisions is important to gain insights in its function as a subcortical relay station and involvement in various pathologies. Yet, detailed histological references of the microanatomy in 3D space are still missing. We therefore aim at providing cytoarchitectonic maps of the medial geniculate body (MGB) and its subdivisions in the BigBrain - a high-resolution 3D-reconstructed histological model of the human brain, as well as probabilistic cytoarchitectonic maps of the MGB and lateral geniculate body (LGB). Therefore, histological sections of ten postmortem brains were studied. Three MGB subdivisions (MGBv, MGBd, MGBm) were identified on every 5th BigBrain section, and a deep-learning based tool was applied to map them on every remaining section. The maps were 3D-reconstructed to show the shape and extent of the MGB and its subdivisions with cellular precision. The LGB and MGB were additionally identified in nine other postmortem brains. Probabilistic cytoarchitectonic maps in the MNI "Colin27" and MNI ICBM152 reference spaces were computed which reveal an overall low interindividual variability in topography and extent. The probabilistic maps were included into the Julich-Brain atlas, and are freely available. They can be linked to other 3D data of human brain organization and serve as an anatomical reference for diagnostic, prognostic and therapeutic neuroimaging studies of healthy brains and patients. Furthermore, the high-resolution MGB BigBrain maps provide a basis for data integration, brain modeling and simulation to bridge the larger scale involvement of thalamocortical and local subcortical circuits.

Deep learning networks reflect cytoarchitectonic features used in brain mapping.

The distribution of neurons in the cortex (cytoarchitecture) differs between cortical areas and constitutes the basis for structural maps of the human brain. Deep learning approaches provide a promising alternative to overcome throughput limitations of currently used cytoarchitectonic mapping methods, but typically lack insight as to what extent they follow cytoarchitectonic principles. We therefore investigated in how far the internal structure of deep convolutional neural networks trained for cytoarchitectonic brain mapping reflect traditional cytoarchitectonic features, and compared them to features of the current grey level index (GLI) profile approach. The networks consisted of a 10-block deep convolutional architecture trained to segment the primary and secondary visual cortex. Filter activations of the networks served to analyse resemblances to traditional cytoarchitectonic features and comparisons to the GLI profile approach. Our analysis revealed resemblances to cellular, laminar- as well as cortical area related cytoarchitectonic features. The networks learned filter activations that reflect the distinct cytoarchitecture of the segmented cortical areas with special regard to their laminar organization and compared well to statistical criteria of the GLI profile approach. These results confirm an incorporation of relevant cytoarchitectonic features in the deep convolutional neural networks and mark them as a valid support for high-throughput cytoarchitectonic mapping workflows.