Neuroanatomy and its Impact on Structural and Functional Imaging (In Memory of Karl Zilles)

This introductory talk aims to provide an overview of the course and introduce ideas on how an understanding of neuroanatomy will impact modern neuroimaging and neuroscience studies. First, I will highlight the necessity of paying attention to neuroanatomy by discussing sulci and gyri in the human brain. Specifically, I will discuss the differences among mammalian species, its relation to functionally-defined brain regions, and its impact on the analysis of neuroimaging datasets. Secondly, I will provide an overview on how the current computational model of neural information processing falls short in representing underlying neuroanatomy, even in the extensively studied visual system. I will introduce various existing models of visual processing, such as the conceptual model of dorsal and ventral visual streams, the more recent hierarchical model of visual processing, and a complex connectivity diagram. Notably, none of these models account for the existence of a relatively large white matter pathway—the vertical occipital fasciculus (VOF)—which connects the dorsal and ventral occipital cortex. By presenting recent evidence regarding the prominence of the VOF in both human and non-human primates, I will discuss the need to revisit the current framework for understanding the model of visual processing in the brain. 

So you think you know how many lobes the brain has? Responses vary between 4-6 lobes among anatomically educated individuals. This variation arises as the correct answer has changed with the evolution of anatomical terminology. This talk will dissect the current debates and their impact on wider society (e.g. Large Language Models are as confused as the scientific community) and cognitive neuroanatomy as a research field. The official nomenclature encompasses staggering 1933 anatomical terms exclusively dedicated to the nervous system. The sheer volume of terms contributes to the intricacy of anatomical language, frequently leading to differences of opinion and discussions. Examining these terms’ etymological roots and significance is crucial for gaining insight into these debates by comprehending their source and influence on our contemporary comprehension of cognitive neuroanatomy, with language serving as a particularly compelling illustration. The anatomy of language is controversial and remains constrained by its adherence to 19th-century models. In this digital age, where much research relies on atlases and automatic methods, it is valuable to reflect on our field's history, embrace recent advancements, and anticipate future challenges. 

The human brain is a marvel of complexity, with its functioning reliant upon a myriad of microstructures and connections. Among the crucial elements of the microstructures, myelin plays a pivotal role. Myelin, which is a fatty substance that envelops nerve fibers, acts as an insulator, facilitating efficient signal transmission within the nervous system. Furthermore, myelin has shown to be critically associated with development, learning, degeneration and various neurological disorders including multiple sclerosis. Therefore, a comprehensive understanding of the spatial distribution and temporal dynamics of myelin is essential for exploring brain structures and functions. Histological studies of myelin distribution in the cortex dates back to 1840, revealing variations in myelin density along cortical depth and among cortical regions. This observation created the concept of “myeloarchitecture”, which refers to intracortical laminar profiles of myelin content across the cortex, similarly to the way cytoarchitecture is used to parcellate the cortex.

In recent years, the field of neuroimaging has witnessed rapid advancements in magnetic resonance imaging techniques, enabling researchers to visualize the in-vivo distribution of myelin in just a few minutes or tens of minutes of scan time. Several methods have been developed, including myelin water imaging (MWI), magnetization transfer (MT) imaging and its variations (MT saturation, and inhomogeneous MT), relaxometric imaging (T1 map and T1 over T2 weighted image ratio), and magnetic susceptibility imaging (quantitative susceptibility mapping, R2* map, and χ-separation). Just like fMRI, where understanding the biophysics behind the technique is important for proper interpretation of the results, it is essential to understand the biophysical foundations of these techniques. In particular, each method has different sensitivity and specificity to myelin and often suffers from various complication factors (e.g., not only myelin but also iron affecting T1 and susceptibility). Therefore, the resulting myelin distribution map inevitably contains errors that vary from one method to another, requiring careful interpretation. No method is considered as a gold standard and prevails. During this presentation, the fundamental principles of these myelin imaging techniques will be described, followed by their technical details, including limitations. After that, validations and key applications of the methods will be included to provide a complete overview of the techniques. Lastly, requirements for an “ideal” myelin imaging technique will be suggested in terms of its need for sensitivity, resolution, etc., for the reliable creation of myelin distribution in future myelin imaging.

Supported by these techniques, several studies have been explored in-vivo myeloarchitecture of the brain, revealing well-known distributions of myelin from histology. I will demonstrate topographical distributions of myelin across cortical regions, illustrating organizational characteristics and also comparing the maps from various myelin imaging techniques against histological results to scrutinize their validity and potential roles for applications including brain parcellation. This examination also unveils the normative distribution of myelin in the healthy brain, offering a chance to compare it with developing, aging and diseased brains. In particular, a few studies that explored such distributions in adolescents and aging brains will be highlighted to understand age-dependent myeloarchitectural changes and their potential impact in brain functions.

In summary, this presentation will provide an in-depth overview of the latest MRI techniques for imaging myelin and myeloarchitecture in the in-vivo human brain. By the conclusion of the talk, attendees will understand the roles of myelin and myelin imaging techniques, currently available myeloachitecture maps, and their applications in exploring neurodevelopment and neurodegenerative diseases. We will navigate through the complex and delicate brain's myeloarchitecture, elucidating its role in our lives, and appreciate how advanced MRI techniques can unlock the secrets it holds. 

Although human cognition has many aspects that can be considered unique, we see the roots of our behaviours in our closest relatives in the animal kingdom. A comparative approach mapping the shared and distinct aspects of brain organisation across species can therefore help us understand the neural machinery that supports brain function. New developments in comparative neuroimaging have paved the way to study a large number of species with great anatomical detail. In particular, diffusion MRI-based reconstruction of brain connections, i.e. tractography, has become a powerful method for comparative neuroanatomy. We know that the general template, or blueprint, of long-range brain connections is shared across mammals. We can therefore reconstruct the path of homologous white matter tracts in different species allowing for a species comparison. In this lecture, I will introduce the concept behind and tools for comparative analysis of diffusion MRI data. The students will get the opportunity to see software demonstrations and will learn about the limitations of the technique. As an example system, I will review new studies on the neurobiology of language and will discuss how comparative analysis approaches have enhanced our understanding of the language connectome. In this context I will also introduce openly accessible databases, for example the Oxford’s Digital Brain Bank and the PRIMatE Data Exchange (PRIME-DE) platform. Furthermore, I will discuss the opportunities that such a comparative framework offers for various research topics, such as evolutionary neuroscience and translational work with animal models. 

The brain evolved around the fundamental homeostatic needs of our organism. Therefore unsurprisingly, sensory afferents representing the physiological condition of the body (or interoception) profoundly influence even our most sophisticated brain functions. Likewise, interoceptive pathologies concur with various neuropsychiatric disorders. Our laboratory examines the interoceptive and autonomic circuits regulating bodily and brain states. In primates, interoceptive afferents form a phylogenetically novel ascending pathway terminating predominantly in the insular cortex. In this talk, I will present architectonics, tract-tracing, electrophysiological and neuroimaging evidence suggesting the existence of (1) a topographic representation of bodily afferents in the dorsal posterior insula or "primary interoceptive cortex", (2) a fine subdivision of the middle insula into distinct architectonic and hodological horizontal "stripes" putatively integrating interoception with poly-sensory inputs, and (3) an ultimate re-representation of contextualized homeostatic states in the ventral anterior insula. I will also show that the morphologically unique, spindle-shaped, von Economo neuron of the ventral anterior insula occupies one specific architectonic area and projects directly back to the primary central relays of interoception as well as to premotor and to pre-ganglionic autonomic centers in the brainstem. Using resting state functional connectivity in neurotypical and coma human subjects, we showed in collaboration that some of these projections may be crucial to associate cortical ‘awareness’ centers with subcortical ‘arousal’ centers, motivating a new series of studies aiming at better appreciating the role of interoception and the insula in brain states. FMRI with concurrent recording or microstimulation in the unique 'von Economo neuron area' confirmed the subcortical projections and unraveled prominent functional relations with high-order cortical areas. These observations support the idea that the ventral anterior insula serves as hub for the simultaneous interoceptive shaping of poly-sensory perceptual experience in an egocentric (‘sentient self’) perspective and the high-order affective regulation of homeostatic bodily states. My talk will illustrate also the advantage of combining fine neuroanatomical methods (architectonics and tract-tracing) with brainwide low spatial resolution fMRI and local high spatial resolution electrophysiology to investigate complex neural systems. It will also conclude on the importance of refined neuroanatomical examination for the comparison of cortical parcellation across species, suggesting the emergence of a higher structural complexity in the human anterior insula, compared to other primate species. 

The cellular anatomy of the human basal ganglia, thalamus and brainstem is generally presented as 2D images in print or digitized formats. 3D reconstruction of brain histology is generally difficult due to a lack of serial sections and technical challenges in reformatting stained 2D slices. Volumetric reconstruction, visualization and annotation of cytoarchitecture can also be challenging in human brain due to large size and high section numbers, which requires intensive computer processing power for 3D visualization and analysis. Unique datasets like the BigBrain allow visualization of the human brain histology at cellular resolution, and advanced software tools allow 3D navigation in any plane and annotation. We will review our recent work that takes advantage of cellular resolution 3-D datasets to annotate nuclei and neighboring tracts of the human basal ganglia, thalamus and brainstem.

Beyond cytoarchitecture, the mammalian brain is characterized by a rich and diverse chemoarchitectural content that defines structure and allows normal function. Compared to non-primate mammals, there is a dearth of detailed knowledge of the chemical anatomy of the primate subcortical brain especially in 3-D format. By way of example, we will present our recent work on creation of a chemo-specific 3D map of the subcortical human brain based on over 2000 serial sections stained with immunochemical markers. We will focus on digital reconstruction and annotation of the dopaminergic system of the brainstem and basal forebrain. To increase utility in brain imaging research these histological and histochemical atlases can be registered to standardized MRI spaces (e.g., The MNI stereotaxic space). As practical clinical applications, these novel atlases can also be used to: a) analyze the MRI location of electrodes in patients who had deep brain stimulation surgery, and b) compare mDA cell types and pathways in healthy brain with similar chemospecific datasets from patients with Parkinson’s disease.