Occipital Intralobar fasciculi and a novel description of three forgotten tracts

The continuously developing field of magnetic resonance imaging (MRI) has made a considerable contribution to the knowledge of brain architecture. It has given shape to a desire to construct a complete map of functional and structural connections. In particular, diffusion MRI paired with tractography has facilitated a non-invasive exploration of structural brain anatomy, which has helped build evidence for the existence of many association, projection and commissural fiber tracts. However, there is still a scarcity in research studies related to intralobar association fibers. The Dejerines’ (two of the most notable neurologists of 19th century France) gave an in-depth description of the intralobar fibers of the occipital lobe. Unfortunately, their exquisite work has since been sparsely referred to in the modern literature. This work gives the first modern description of many of the occipital intralobar lobe fibers described by the Dejerines. We perform a virtual dissection and reconstruct the tracts using diffusion MRI tractography. The virtual dissection is guided by the Dejerines’ treatise, Anatomie des Centres Nerveux. As an accompaniment to this article, the authors provided a French-to-English translation of the treatise portion concerning intra-occipital fiber bundles. This text provides the original description of five intralober occipital tracts, namely: the stratum calcarinum, the stratum proprium cunei, the vertical occipital fasciculus of Wernicke, the transverse fasciculus of the cuneus (TFC) and the transverse fasciculus of the lingual lobe of Vialet. It was possible to reconstruct all these tracts except for the TFC – possibly because its trajectory intermingles with a large amount of other fiber bundles. For completeness, the recently described sledge runner fasciculus, although not one of the Dejerine tracts, was identified and successfully reconstructed.


INTRODUCTION
The advent of novel in vivo magnetic resonance imaging (MRI) techniques has heightened the desire to construct a comprehensive atlas of the human brain connective networkthe so-called human connectome Shibata et al 2015;Sotiropoulos and Zalesky 2019). In humans, diffusion MRI paired with tractography permits a non-invasive three-dimensional reconstruction of large-scale white matter system organisation, which can then be compiled into an atlas (Wakana et al 2004)  Despite the rapid advancement in the technology and the connectome mapping techniques available, there is a paucity in research related to short-range association fibers in the occipital lobe. The initial focus of this new field was instead geared towards the more prominent, and seemingly more physiologically crucial, long-range association fibers. An additional issue with MRI-based tractography studies is that the findings tend to remain speculative, devoid of tangible experimental evidence and ground truth, unlike those of axonal tracer studies (Donahue et al 2016).
To partially overcome the limitations inherent in diffusion tractography, contemporary researchers can guide their findings through post-mortem dissection studies, by consulting the tract tracer literature, or by consulting historical neuroanatomy texts.
The 19 th -century neurologists were masterfully skilled in gross dissection and fastidious in the documentation of their composite observations. As such, their findings remain relevant and still permeate today's literature. Numerous diffusion tractography studies have successfully used classical dissections to interpret imaging results and identify false positives or false negatives, and have been successful in building on the groundwork made by their predecessors (Schmahmann and Pandya 2006;Forkel et al 2014;Vergani et al 2014;Yeatman et al 2014). Pioneering texts have also often been an initial source of inspiration for explorative neuroanatomical studies (Forkel et al 2014;Yeatman et al 2014). Indeed, the authors had previously studied dissections by 19 th -century neurologists Joseph and Augusta Dejerine with a view to enriching the field's knowledge and perspectives on long-range association fiber anatomy (Bajada et al 2017). It showcased the pertinence of these historical studies and prompted further examination into the Dejerines' other work on association short-range occipital fibers.
The 1900s was a century particularly rich in pivotal studies and debates between different schools of thoughts. One of the biggest divisions across the neuroanatomical field towards the end of the 19 th century was the definition and categorisation of association fibers. Meynert Dejerine (1849Dejerine ( -1917 and his wife, Augusta Dejerine-Klumpke (1859-1927 frequently collaborated and are both individually celebrated as two of France's most renowned 19 th -century neurologists (Bogousslavsky 2005;Shoja and Tubbs 2007;Bajada et al 2017). Together, they pioneered anatomofunctional studies of the cerebral association fiber tracts and challenged the period's paradigm of brain organisation (Shoja and Tubbs 2007;Krestel et al 2013;Bajada et al 2017 (Chou et al 2009;. Of particular interest to the authors, the Dejerines dedicated a portion of the first volume to occipital intralobar association fibers. The Dejerines observed five fiber bundles specific to the occipital lobe: the stratum calcarinum, the vertical occipital fasciculus (VOF) of Wernicke, the transverse bundle of the lingual lobule of Vialet, the transverse fasciculus of the cuneus and the stratum proprium cunei. At the time, intralobar white matter fibers were just being discoveredthese were as yet poorly understood and surrounded by controversies due to inconsistent classification and nomenclature (Yeatman et al 2014;Mandonnet et al 2018), leading to much of the Dejerines' work on their trajectory being overlooked.
Of late, the VOF, the largest of the occipital intralobar fibers dissected by the Dejerines, has been the focus of an increasing number of papers (Weiner et al 2017;Güngör et al 2017;Oishi et al 2018;Briggs et al 2018;Panesar et al 2019). As of December 2019, a PubMed search generated thirty hits relating to the "vertical occipital fasciculus" when used as a key term. In comparison, the other intraoccipital tracts generate no results and remain overlooked, mostly being listed in passing. An article by our group revisited the Dejerines' work and provided access to the first English translation of the portion of the Dejerines' book related to the long-range association fibers of the human brain (Bajada et al 2017). Here, we provide the first complete translation of the description by the Dejerines of the occipital originating short-range fibers trajectories, from the original French taken from Anatomie des Centres Nerveux (1895 p.780-784) (supplementary material 1 and 2). The aim of this study is to use the observations made by the Dejerines for guidance through a virtual dissection of the intralobar tracts of the occipital lobe, based on in vivo diffusion imaging data. The virtual dissections enable us, to have a better discern the threedimensional relations of these once forgotten tracts.

MATERIAL AND METHOD
The study consisted of two principal methods: the first being the translation from French into English of the Dejerines' classical study of occipital intralobar fibers. This was followed by a diffusion weighted MRI (dMRI) based virtual reconstruction of the Dejerines' dissection in healthy subjects.

TRANSLATION
The masterful description of intra-occipital fiber tracts in Chapter 5: pp.780-4 by J.J. Dejerine and A. Dejerine-Klumpke in Anatomie Centre Nerveux;Volume I (1895) was fully translated (MB) from its original 19 th -century French to more universally accessible English and reviewed for its accuracy by the experts in this field (CB, MTS

IMAGE ACQUISITION
The full details of image acquisition can be found in the following articles (

RESULTS
Using CSD tractography, this study successfully identified and mapped four of the five occipital intralobar white matter tracts observed by the Dejerines. The translation of the Dejerines' work on these fibers, from 19 th -century French to accessible English

STRATUM CALCARINUM (SC)
The stratum calcarinum (SC) consists of several U-shaped fibers that connect the two banks of the calcarine fissure, a major landmark that medially divides each occipital lobe into the wedge-shaped cuneus superiorly, and the lingual lobule inferiorly, and represents the boundaries of the primary visual cortex. This group of U-shaped fibers extends all along the calcarine fissure from the occipital pole to its junction with the parieto-occipital fissure at anterior border of the occipital lobe ( Fig.   2). The SC is an unusually large U-fiber system because it directly connects adjacent gyri across a long and deep fissureperhaps the reason the Dejerines devoted considerable attention to it, describing it alongside other intralobar fibers.

TRANSVERSE FASCICULUS OF THE LINGUAL LOBE OF VIALET (TFV)
A tract resembling the many aspects of the Dejerines' description of the transverse fasciculus of the lingual lobe of Vialet (TFV) was successfully isolated (Fig.4). The Dejerines depicted the TFV as a mirror counterpart to the transverse fasciculus of the cuneus, as it connects the inferior gyrus of the calcarine fissure to the inferolateral aspects of the occipital lobe, akin to the way in which the transverse fasciculus of the cuneus connects the fissure to the external surface of the cuneus.
In contrast, the stratum proprium cunei connects the fissure to the medial surface of the hemisphere and does not appear to have an inferior lingual gyrus equivalent.
Interestingly, the TFV origin can be found all along the calcarine fissure alongside the stratum proprium cunei.
The TFV has a morphology that resembles an extended U-fiber; the tract travels laterally away from the fissure prior to adopting an intense anterior-oblique trajectory that sharply curves back on itself and continues in a posterior-oblique course toward the lateral convexity of the occipital lobe. Due to the squatness of the lingual lobe, the tract appears quite flat and two ends of the fiber tract are nearly on the same coronal plane, with the lateral terminations slightly inferior to its calcarine fissure origin. The TFV starts superiorly and medially to the inferior longitudinal fasciculus and occipital projections of the inferior fronto-occipital fasciculus, and crosses laterally over these two major association fiber bundles, before descending just laterally to them.

VERTICAL OCCIPITAL FASCICULUS (VOF)
Using the Dejerines' classical description, we identified the vertical occipital fasciculus (VOF) (Fig.5A), and it aligns well with modern tractography findings. Of the five occipital tracts the Dejerines observed, the VOF is the longest and most prominent, and consequently has been the focus of many more studies. Often described as a short, vertical, and 'sheet-like' bundle of fibers, the VOF "connects the superior aspect of occipital lobe to its inferior surface" (

DISCUSSION
The anatomical described in the present study suggests that the occipital lobe has at least five intrinsic white matter structures that fall between the long association fibers and the very short U-shaped fibers. These are: ). It is important to make these distinctions, for both the VOF and other intralobar fibers, in order to generate a truly representative mapthat is, one that does not include pathways that appear longer or wider than they truly are because they artificially contain subsections of adjacent tracts. Knowing the extent of a distinct white matter system is useful for future work concerning functional specialization and pathology. It has been suggested that certain U-fibers, situated between the VOF and other long association pathways, form artificial connections, or U-fiber 'bridges', between the pathways that can make the two tracts look like one larger tract (Panesar et al 2019). The intermingling of U-fibers is known to obstruct the visualisation and interpretation of the deeper underlying subcortical tracts in both fiber tractography and dissections (Reveley et al 2015).
The difficulty that arises when studying the work of the Dejerines, like other 19 thcentury anatomists, relates to translating their definition of structures into the current school of thought. Inconsistent landmarks, nomenclatures and methods were used to define and visualise structures. This discordance makes it difficult to reference a classic text and persists to the present day. Nevertheless, the congruency between the Dejerines' dissection and modern imaging suggests that the other tracts they described should also be successfully visualised in vivo.

STRATUM CALCARINUM (SC)
In vivo and in vitro dissections of the calcarine fissure reveal a large continuous bundle of U-fibers, named the stratum calcarinum (SC), connecting the upper and lower edges of calcarine cortex (Fig.2). This is the first tractography reconstruction of the SC as U-fiber studies are infrequent. U-fibers typically do not form a clear pathway or large fiber system. The Dejerines chose to omit the SC from the portion of their book that concerns U-fibers, in favour of discussing it alongside the association fibers of the occipital lobe to highlight the prominence and size of the SC.
The Dejerines emphasised that it is not a true association fiber or a single entity.
Instead, it represents a series of tracts forming an oversized U-fiber system that curves around the whole calcarine fissure and is specific to the primary visual cortex.
It extends from the posterior end of the occipital horn to the intersection with parietooccipital fissure and circumnavigates the calcar avis in its full extension. The SC intermingles with the other association fibers like the transverse fasciculus of Vialet and stratum proprium cunei, which link the primary visual cortex with the convexity of the occipital lobe.
The SC system forms an atypical bundle as it consists of a long and deep layer underlying a short and more superficial layer. The superficial SC layer, as can be seen from the results ( fig. 2C), has short projections that tightly carpet the floor of the calcarine fissure and join the superior and inferior gyri that directly border the fissure.
This gives the layer the characteristic 'U' appearance. The second underlying layer has wider projections and a weaker U shape. Its longer fibers connect the medial aspect of the central cuneus to the inferior-medial aspect of the lingual lobe.
The calcarine fissure receives several white matter bundle pathways that share similar orientations at the level of the fissure, making it particularly difficult to differentiate them. Campbell (1905), postulated that the majority of the fibers labelled by the Dejerines, such as the SC, were actually terminal fibers of the optic radiation.

STRATUM PROPRIUM CUNEI (SPC)
The SPC is a short and seemingly inconsequential tract that rises in the sagittal plane from the calcarine fissure at the medial border of the hemisphere ( fig.3). Since the first reports by Sachs (Sachs and Wernicke 1892), it has only been vaguely mentioned in passing (Campbell 1905;Von Bonin et al 1942;Greenblatt 1973 (Forkel et al 2014), which was later shown to be inaccurate by more focused tractography studies. Vergani et al. (2014), conducted a thorough post-mortem exploration of the intralobar fibers of the occipital lobe, in an effort to replicate the work of Sachs (Sachs and Wernicke 1892). Interestingly, like Sachs, the authors did not mention a tract that matches the morphology of the TFV as described later by Vialet, the Dejerines, and as identified in this study. A recently resurfaced photograph of a Weigert-stained histological slide of the macaque occipital lobe by von Bonin et al. (von Bonin et al., 1942) clearly depicts the vertical occipital fasciculus and stratum calcarinum, and somewhat less clearly, the TFV and transverse fasciculus of the cuneus named the fasciculus transversus lingualis and fasciculus transversus cunei, respectively . The brevity of the description makes it hard to ascertain in greater detail the course of the TFV in the macaque.

THE TRANSVERSE FASCICULUS OF THE CUNEUS (TFC)
Our study did not identify a consistent tract in either hemisphere that matches any previous descriptions. Yet, the Dejerines remarked that the calcarine cortex is attached to the cuneus and first occipital convolution through both the stratum proprium cunei and TFC and this observation of the TFC was also made by (Sachs and Wernicke 1892;Vialet 1893;Campbell 1905;Von Bonin et al 1942;Greenblatt 1973;Vergani et al 2014). The Dejerines, with reference to Sachs' work, described the TFC as having a trajectory that "runs slightly obliquely, anteriorly and laterally, and radiate[s] in the superior parietal lobule and in the angular gyrus" (Dejerine, 1895 pg.781 (Vergani et al., 2014).
The two opposing origins along the calcarine fissure of the transverse fasciculus of Vialet and TFC, and the mirroring trajectories described by the Dejerines, imply the possibility of complementary functions related to the upper and lower quadrants of the visual field, respectively, and integration of visual stimuli to the accessory visual integration area and the rest of the cerebrum (Campbell 1905;Greenblatt 1973). It follows that if the transverse fasciculus of Vialet exists to connect the areas of the cortex involved in the superior vision fields with distal cortical regions, a similar subcortical system would need to be in place for the inferior vision fields.

THE SLEDGE RUNNER FASCICULUS (SRF)
The SRF had not been described by Sachs, nor later by the Dejerines or other neuroanatomists until Vergani and colleagues (Vergani et al 2014). The novel tract was initially named the 'sledge-runner fasciculus' to highlight its undulating shape due to its anteriorly projecting curve at the level of the calcarine fissure (Vergani et al 2014). Later, in a focused tractography study, the SFR was alternatively named as the medial occipital longitudinal tract (MOLT) to better reflect its anatomy (Beyh et al 2017). We continue to use the term SRF as it is the more widely known term. The fibers. The first two likely led to artefacts and unclear structures, and the latter incorporated a bias into the interpretation of their findings. There are therefore limits to using historical dissection work to guide new findings. This is especially the case since different neuroanatomists omitted certain tracts from their atlases that they judged, rightly or wrongly, to be inaccurate. Historical work has proven to be a useful starting point for reviving unexploited concepts. It also serves as a strong guideline in helping to determine the potential value of a study. Regardless of the above, they do not circumvent the need for subsequent validation from the modern post-mortem dissection literature.
Tractography has generated findings shrouded in controversies, many dating from the 19 th century that are still unresolved. Yet, great strides have been made in acknowledging and reducing biases. In this study we have shown that there is a significant concordance between the original anatomical work of the Dejerines and the recently well-defined VOF, giving us confidence to extrapolate trust in our findings of the other lesser-known tracts. The results indicate gaps in current knowledge and encourage future studies based on this new perspective of what the association fibers of the occipital lobe entail.