ABSTRACT
Accurate neuroanatomical identification relies on understanding the composition and arrangement of the neurons and fibers that make up each brain structure. Numerous parcellations have been based on differences in cytoarchitecture, myeloarchitecture, chemoarchitecture, and pigmentoarchitecture to create architectonic brain maps (Campbell, 1905; Elliot Smith, 1907; Brodmann, 1909; Vogt and Vogt, 1919; von Economo and Koskinas, 1925; Flechsig, 1889; Vogt, 1910; Dietl et al., 1987; Zilles et al., 1988; Burkhalter and Bernardo, 1989; Campbell and Morrison, 1989; Jansen et al., 1989; Clarke, 1994; Hendry et al., 1994; Tootell and Taylor, 1995; Zilles et al., 1995; Braak and Braak, 1997; Braak, 1979; Ding et al., 2016). For this chapter, we will focus on the cyto- and myeloarchitecture of the human brain. The cytoarchitecture of the human brain is the study of the arrangement of cells within the tissue, while the myeloarchitecture is the study of the arrangement and density of myelin that surrounds the axons of the neurons in the cerebral cortex. Cytoarchitecture encompasses the nature of neurons themselves, their type, size, and morphology, as well as their organization into various cortical areas. In the cortex, the neurons are arranged in six layers, layer I being the closest to the pia mater and layer VI being adjacent to the white matter. The cerebral cortex of the human brain has been parcellated into cortical areas that exhibit variations – some slight and some considerable – in cytoarchitecture. The regions, frequently called Brodmann areas (BA), are named after Korbinian Brodmann, a German neurologist who published his work on brain parcellation based on cortical cytoarchitectonics in 1909. Each of these cortical regions also exhibits its own myeloarchitecture, as shown by Paul Emil Flechsig, a German neuroanatomist, pathologist, and psychiatrist (Flechsig 1889). The thickness and location of the layers in the gray matter vary between adjacent regions. The composition, such as the type, morphology, and density of neurons, as well as the myelin fiber density, vary across different regions. The most common stain to visualize the cortical architecture is a Nissl stain for neurons with a viable cytoplasm (Amunts et al., 1999; Augustinack et al., 2005) and to visualize myelinated fibers, a Gallyas stain (Pistorio et al., 2006). However, histology is a labor-intensive technique. It requires first sectioning the tissue with a vibratome or a cryomicrotome, then mounting the sections onto glass slides, which are dried, histologically stained, and coverslipped for preservation. This process is likely to introduce irremediable distortions at almost every step. Some of the distortions and artifacts from histology are demonstrated in Section 18.5.4 Advantages and Limitations. Full 3D reconstruction of the human brain from histological sections (stained for intact neurons) has been published by Amunts et al. (2013). Manual and automatic repairs on the damaged slices were performed as well as non-linear registration of the coronal slices to the MRI to reconstruct the entire volume. Even after this labor-intensive work, the BigBrain still has registration artifacts that are apparent in the other orientations (sagittal and axial). Registration between histology slices is difficult (Arsigny et al., 2005), even with the aid of intermediary images of the undistorted tissue, such as the blockface images or the MRI (Osechinskiy and Kruggel, 2010). Those distortions at the microscopic level can be avoided by imaging the blockface directly, prior to cutting (Odgaard et al., 1990). In contrast to histology, OCT probes the superficial part of a tissue block (a few hundred microns to a millimeter in depth depending on the tissue and the fixation), prior to any sectioning. Thus, OCT can be used to reconstruct brain samples of several cubic centimeters without distortion and with contrasts comparable to histology, as will be shown in the following sections, facilitating the reconstruction of a micron-resolution volume.
