Received: 5 December 2021 Revised: 14 March 2022 Accepted: 9 May 2022 DOI: 10.1002/hbm.25964 RESEARCH ARTICLE Sulci 3D mapping from human cranial endocasts: A powerful tool to study hominin brain evolution Edwin John de Jager1,2 | Laurent Risser3 | Muriel Mescam4 | Caroline Fonta4 | 1,5,6 Amélie Beaudet 1 Department of Archaeology, University of Cambridge, Cambridge, UK Abstract 2 Department of Anatomy, Faculty of Health Key questions in paleoneurology concern the timing and emergence of derived cere- Sciences, University of Pretoria, Pretoria, bral features within the human lineage. Endocasts are replicas of the internal table of South Africa 3 Institute de Mathématiques de Toulouse, the bony braincase that are widely used in paleoneurology as a proxy for Université de Toulouse, UPS, Toulouse, France reconstructing a timeline for hominin brain evolution in the fossil record. The accu- 4 Centre de Recherche Cerveau et Cognition rate identification of cerebral sulci imprints in endocasts is critical for assessing the (CerCo), CNRS, Université de Toulouse, UPS, Toulouse, France topographic extension and structural organisation of cortical regions in fossil 5 School of Geography, Archaeology and hominins. High-resolution imaging techniques combined with established methods Environmental Studies, University of the Witwatersrand, Johannesburg, South Africa based on population-specific brain atlases offer new opportunities for tracking 6 Institut Català de Paleontologia Miquel detailed endocranial characteristics. This study provides the first documentation of Crusafont, Universitat Autònoma de sulcal pattern imprints from the superolateral surface of the cerebrum using a Barcelona, Barcelona, Spain population-based atlas technique on extant human endocasts. Human crania from Correspondence the Pretoria Bone Collection (South Africa) were scanned using micro-CT. Endocasts Edwin John de Jager, Department of Archaeology, University of Cambridge, were virtually extracted, and sulci were automatically detected and manually labelled. Cambridge, UK. A density map method was applied to project all the labels onto an averaged endo- Email: edwinj.dj@gmail.com cast to visualise the mean distribution of each identified sulcal imprint. This method Funding information allowed for the visualisation of inter-individual variation of sulcal imprints, for exam- South Africa/France (PROTEA) Joint Research Programme, Grant/Award Number: 129923; ple, frontal lobe sulci, correlating with previous brain-MRI studies and for the first University of Cambridge Harding Distinguished time the extensive overlapping of imprints in historically debated areas of the endo- Postgraduate Scholars Programme cast (e.g. occipital lobe). In providing an innovative, non-invasive, observer- independent method to investigate human endocranial structural organisation, our analytical protocol introduces a promising perspective for future research in paleoneurology and for discussing critical hypotheses on the evolution of cognitive abilities among hominins. KEYWORDS atlas, density map, human cortex, micro-CT, sulcal pattern, virtual endocasts This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2022 The Authors. Human Brain Mapping published by Wiley Periodicals LLC. Hum Brain Mapp. 2022;43:4433–4443. wileyonlinelibrary.com/journal/hbm 4433 4434 de JAGER ET AL. 1 | I N T RO DU CT I O N suggest that correct interpretation of details on the endocranial sur- face almost always relies on the neuroanatomical knowledge of the Since brains are not preserved in the fossil record, paleoneurologists  n, 2013). An additional, less explored researcher (Zollikofer & De Leo are presented with the challenge of relying on the imprints preserved (albeit no less important) factor that may bias our interpretations of on the inner surface of the braincase. These imprints are formed the fossil record is intra-specific sulcal pattern variation. through an intricate balance of tensions and pressures during morpho- Owing to the early works of Connolly (1950), researchers were genesis of the brain and braincase (Moss & Young, 1960). For this rea- able to study the extensive sulcal variations of the human brain. son, researchers make use of endocasts, a moulded cast of the inner Brain atlases that followed, such as those by Ono et al. (1990) as well surface of the braincase, to extract the negative space originally occu- as Duvernoy and Bourgouin (1999), provided further evidence of the pied by the brain and to infer conclusions on brain size, shape, and variation of the extant human sulcal pattern. The integration of med- general surface morphology, such as sulcal patterns and meningeal ical imaging techniques into neurosciences opened new perspectives vascular distribution (Falk, 1980a, 1987, 2014; Holloway, 1978, 2009; on how to quantify this variation. In particular, the substantial intra- Holloway, Broadfield, & Yuan, 2004). In particular, sulcal imprints have specific variability in extant human brains has long been a challenge been the focus of much interest in paleoneurology (Falk, 2014; for neurosurgeons when trying to delineate cortical features during Holloway, 1981; Le Gros Clark et al., 1936). Sulci form relatively early keyhole surgery (Ribas, 2010). To solve this problem, researchers during foetal development; the first of which to form is the longitudi- have created population-specific brain atlases (e.g. BrainVISA; publi- nal fissure (gestational week: 8–22), separating the two cerebral hemi- shed by Rivière et al., 2009) that provide a method for data from spheres. Primary sulci, namely: the Sylvian fissure, as well as the multiple subjects and modalities to be compared and analysed (Toga cingulate-, parieto-occipital- and calcarine sulci (gestational week: 14– et al., 2006). These atlases retain information pertaining to intrapop- 16); the central- and superior temporal sulci (gestational week: 20– ulation brain variations, proving useful in the detection of pathology 24); and the superior frontal-, postcentral- and intraparietal sulci (ges- in patients and assistive in functional image analysis (Thompson tational week: 25–26), are responsible for the subdivision of the hemi- et al., 2000; Toga & Thompson, 2001), and furthermore, revealing spheres into the frontal, parietal, occipital and temporal lobes, previously unreported fascinating aspects of the human brain respectively; which were named according to their corresponding (Glasser et al., 2016). bones of the skull. Secondary sulci (gestational week: 30–35) and ter- Toga and Thompson (2001) described three methods that form tiary sulci (gestational week: 36) follow in development well into the the foundation of population-based atlases. First, the popular label- postnatal period, despite volumetric expansion of the brain slowing by based approach (also known as statistical/probabilistic anatomy maps) the age of 6 years (Stiles & Jernigan, 2010). The intimate morphologi- involves large volumes of brain data (cortical structures) being manu- cal relationship existing between the growing brain and the surround- ally segmented after the data set is aligned in a stereotaxic space. Sec- ing cranial bones (Moss & Young, 1960) results in these structures ond, the deformation-based approach, (based on a warping algorithm), being imprinted onto the cranial inner surface very early in one indi- is used to calculate a deformation map of a scan, based on a reference vidual's life; these imprints can thus be observed in immature and scan optimised by the input atlas; thereby matching cortical structures mature fossil specimens (e.g. Falk, 1980b; Lockwood & Tobias, 1999). from one scan to their corresponding structure in another (reference) However, these endocasts have limitations. Firstly, information is scan (Bookstein, 1989). In more complex cases, magnitude and direc- limited to the external cortical surface, and nothing can be inferred tional biases of variations can be determined. Furthermore, assess- about the subcortical structures (Holloway, Broadfield, & Yuan, 2004). ments of the severity of structural variants outside the normal range Second, fossil crania are often severely damaged and incomplete or can be made to determine if pathology is present. Finally, the density- distorted due to taphonomic processes (Beaudet & Gilissen, 2018; based approach focuses on producing generalised representations of Neubauer, 2014). Other limitations reported are related to the pres- the anatomy by averaging the intensity of every voxel from multiple ence of intracranial components (e.g. arterial supply of the brain, cere- 3D images (e.g. MRI) on a voxel-by-voxel basis and aligning the scans brospinal fluid, and meningeal membranes) that may blur cerebral into a stereotaxic space to create an average image database details on the endocranial surface (Neubauer, 2014). That being said, (Bookstein, 1989; Evans et al., 1994; Thompson et al., 2000; Dumoncel et al. (2021) recently reported that there is a good correla- Thompson & Toga, 1997, 1999; Toga & Thompson, 2001). tion between the endocast (CT) and the brain (MRI), of the same indi- Within this context, we propose that a better understanding of vidual, when considering shape and sulcal patterns of the frontal the intra-specific variation of the sulcal pattern in extant human endo- lobes, temporal lobes, occipital lobes and inferior portion of the parie- casts would significantly improve interpretation of the fossil record by tal lobes. They further reported that organisations and morphology of avoiding confusion in identifying the imprints. Here we apply the con- the superior part of the brain are questionable (Dumoncel cept of population-based brain atlases, as developed in neuroscience. et al., 2021). Most of the uncertainties in human paleoneurology have Specifically, we apply the density-based approach (ideal for aligning resulted from these limitations, which have contributed to long- images on a voxel-by-voxel basis due to the high interindividual varia- standing debates. For example, the identification of the lunate sulcus tion inherent to our sample) to characterise the variation of the sulcal in the endocast of the Taung child (Australopithecus africanus; pattern imprints in a sample of extant human endocasts. This study Falk, 1980b; Holloway, 1981). In this context, it may be tempting to ultimately provides a crucial population level reference from which to de JAGER ET AL. 4435 guide subsequent analyses of fossil hominin endocasts, with the vision Ci,j as the list of the labelled curves that represent the sulcal imprints of reducing bias in future sulcal imprint identification endeavours. on Ei. The index j represents a specific label that corresponds to each sulcal imprint, with j = 1,…,p. The name of this label was further anno- tated as Lj. Note that labelled curve Ci,j may consist of multiple 2 | MATERIALS AND METHODS segments. Finally, the density maps were estimated based on the spatial dis- 2.1 | Materials tribution of the sulcal imprints. This pipeline is summarised in Figure 1, and the different steps are detailed below. In this paper, we used the sample published in de Jager et al. (2019). This sample includes virtual endocasts extracted from 20 extant human crania with an equal proportion of males and females aged 2.2.2 | Average endocast surface 30–80 years. No pathology affecting the cranial bones was visible. All crania are curated at the Pretoria Bone Collection (see L'Abbé All the meshes representing the endocasts and curve sets rep- et al., 2005). resenting the sulcal imprints (both in .vtk format) were rigidly aligned and scaled during the extraction and detection process. To calculate the average shape Eav of the whole endocast sample set, a void 2.1.1 | Detection and identification of sulcal 200  200  200 voxels 3D image (volume file) I(Mi) was first gener- imprints ated for each mesh Mi. The correspondence between the mesh's (.vtk format) coordinates and the 3D image coordinates was defined so Dry human crania were scanned at the South African Nuclear Energy that all meshes were represented as 3D images with a margin of about Corporation (Necsa), Pelindaba, South Africa using micro-focus X-ray 10 voxels (10 mm). A value of one was assigned to the voxels in the tomography (micro-CT; Nikon XTH 225 ST) at a spatial resolution image corresponding to each node on the corresponding mesh. Thus, ranging from 94–123 μm (isometric voxel size). Virtual endocasts 3 the meshes were densely sampled in the 3D images. The resulting 3D were generated by using Endex, an automatic segmentation software images were then smoothed using a Gaussian filter with a standard (Subsol et al., 2010; available at: https://perso.liris.cnrs.fr/gilles. deviation of five voxels (5 mm). We denote I_Mi as the densified and gesquiere/wiki/doku.php?id=endex). A semi-automatic detection smoothed endocast surface in the future image domain of the average method was used to automatically identify the sulcal imprints on the shape A_M. surface of the endocasts. The algorithm automatically detects topo- Once all the 3D images I_Mi were computed, their average shape graphical variations, for example, crest lines on meshes to indicate was defined as the barycentre of these images in a specific cortical relief in endocasts (Yoshizawa et al., 2007, 2008; available Riemannian-shaped space following the method by Fiot et al. (2014) from: http://www2.riken.jp/brict/Yoshizawa/Research/Crest.html). and Vialard, Risser, Rueckert, and Holm (2012). A 3D image was ran- By using a custom script (available from: https://gitlab.com/ domly picked out of the n images (I_Mi) to be the initial guess average jeandumoncel/curve-editor) in Matlab R2013a (Mathworks), detected image. It was then iteratively deformed towards the Karcher mean surface features of interest (e.g. imprints of sulci) were manually iden- (Karcher, 1977), considered (as) the barycentre of the n images (I_Mi) tified and labelled by reviewing brain atlases of Connolly (1950) and in a Riemannian space, of all the images of the entire sample. The Ono et al. (1990), and endocast atlases (Beaudet et al., 2016, 2019; main advantage of this method was that it fully preserved the topol- refer to de Jager et al., 2019 for more details). All unwanted artefacts ogy of the cranial structures. In practice, this iterative method was detected by the algorithm (e.g. imprints formed by meningeal arteries performed as follows: We denote A_Mcur as the estimated average and cranial sutures) were systematically removed. image at the current iteration. At the (very) beginning of the iterative process, A_Mcur is initiated as being equal to the randomly picked image among the n images (I_Mi). Diffeomorphic registration of A_Mcur 2.2 | Density maps on the different images I_Mi was then performed using the method of Vialard, Risser, Rueckert, and Cotter (2012), which resulted in n scalar 2.2.1 | Overview fields encoding the deformation and denoted initial momenta mi. At each iteration, the current average A_Mcur is then updated by being Here we briefly summarise the approach we developed to compute deformed using the information in the average momentum mcur = the anatomical variability of sulcal imprints based on a sample of mean(m1,…, mn). This process is repeated until A_Mcur does not present n endocasts (n = 20). We denoted each image of an endocast as Ei, any significant deformation. As shown in Vialard, Risser, Rueckert, and where i depicts values of 1 to n. We first computed a standardised Cotter (2012), this process typically converges after about 5–10 itera- surface of the endocast by using the extracted meshes. Then, we pro- tions for skulls. Remark that a Karcher mean is defined with respect to jected all the detected and identified sulcal imprints to this a Riemannian Kernal. This kernel models the level of smoothness standardised surface. We denoted Mi as the mesh associated with Ei, allowed when moving A_Mcur to its next estimation. As a conse- which specifically represents the endocast. We additionally denoted quence, this kernel models the scale at which the average shape A_M 4436 de JAGER ET AL. F I G U R E 1 Workflow of steps followed to create density maps. Blue: Sulcus detection and labelling procedure; green: Creating average endocast; Orange: Creating density maps for each sulcal imprint is computed. In our study, we used a Gaussian kernel with a standard too much smoothing (which would decrease the localisation power of deviation of 10 mm. This means comparing shapes having a much this method as well as mix information from different sulci) and not larger scale than 10 mm is meaningless (estimations will be unstable), enough smoothing for which the density maps would start only rep- and that the structures having a scale lower than 10 mm are consid- resenting each individual sulcus. Its intensities were finally normalised ered as details. We considered this kernel as adapted to compute so that their sum was one. Each resulting density map I_Cj subse- average endocasts. It however cannot take into account shape details quently represented the spatial variability of each observed sulcus. having a size lower than 10 mm. The average shape A_M is finally The variations were documented up to two standard deviations. equal to A_Mcur at convergence. 3 | RE SU LT S 2.2.3 | Sulcal imprint variability estimation In general, all sulci (n = 19) from the superolateral surface of the brain Although computing the average image A_M presents a relatively lim- were identified on 20 endocasts. On average, 73% and 65% of frontal ited interest in itself, it interestingly allows to compute a set of sulci, 42% and 35% of parietal sulci, 88% and 93% of temporal sulci n invertible displacement fields (deformation) ϕi which spatially map all and 51% and 40% of occipital sulci were identified from the left hemi- features of the Ei to this common template. Each curve set Ci,j was spheres (LHs) and right hemispheres (RHs) respectively. Additionally, then spatially projected at the surface of the averaged endocast using the Sylvian fissure was clearly identified in almost all endocasts ϕi. This projected set was denoted as P(Ci,j). Considering a specific sul- (LHs = 95%; RHs = 85%) (de Jager et al., 2019; Table S1). cal label Lj out of all pre-defined labels, all known curves P(Ci,j) were The following results of sulcal impressions are listed in a general represented in a single image sampled in the domain of Eav (i.e. an anterior to posterior order of their appearance on the brain. The image of 2003 voxels). Note that this image only describes the location impressions of fronto-marginal and fronto-orbital sulci could be seen of the curves associated with label Lj. We denote this image as I_Cj this showing centralised distributions along the caudal margin of the fron- image. In order to represent the spatial variability of the sulci labelled tal lobe (Figure 2a,b). Similar results can be seen for the superior fron- as Lj on the standardised endocast surfaces, we smoothed each image tal sulcus, which showed a localised distribution more rostrally along I_Cj with a Gaussian filter with the standard deviation of five voxels the medial margin of the frontal lobe (Figure 2c). Furthermore, the (i.e. 5 mm). This smoothing level appeared as a good trade-off between middle frontal sulcus and the inferior frontal sulcus showed higher de JAGER ET AL. 4437 F I G U R E 2 Density map indicating the distribution of sulcal imprints around the mean in left and right views. Red indicates the mean distribution and blue indicates mean + 2 standard deviations. (a) Fronto-marginal sulcus; (b) Fronto-orbital sulcus; (c) Superior frontal sulcus; (d) Middle frontal sulcus; (e) Inferior frontal sulcus; (f) Precentral sulcus; (g) Central sulcus; (h) Postcentral sulcus variation rates (larger green and blue areas) than the superior frontal precentral sulcus clearly showed a localised distribution (indicated in sulcus (Figure 2d,e). When comparing the distribution between the red) along the caudal margin of the frontal lobe and more outliers LHs and the RHs, the superior frontal sulcus shows a more rostrally towards the superior margin of the hemisphere particularly on the LHs. centred distribution in the LHs compared to the elongated red area on Similarly, the vertically orientated left central sulcus showed a more the RHs (Figure 2c) Additionally, the characteristic caudal and rostral centralised red distribution located in the middle of the hemisphere just centred distribution of the middle frontal sulcus is more prominent on above the temporal lobe, marking the border between the frontal and the RHs and more extreme variations, indicated as green on the den- parietal lobes and more variability towards the superior margin sity maps (Figure 2d). (Figure 2g). The postcentral sulcus showed highly random distributions Scattered distributions were observed for the precentral sulcus, on the LHs and a clear centralised distribution on the RHs (Figure 2h). central sulcus, postcentral sulcus, and intraparietal sulcus, particularly The Sylvian fissure shows a clear horizontal distribution on the on the LHs (Figure 2f,g,h and Figure 3g). The distribution of the right lateral surface of the endocast, separating the temporal lobe from the 4438 de JAGER ET AL. F I G U R E 3 Density map indicating the distribution of sulcal imprints around the mean in left and right views. Red indicates the mean distribution and blue indicates mean + 2 standard deviations. (a) Anterior horizontal ramus; (b) Ascending ramus; (c) Diagonal sulcus; (d) Sylvian fissure; (e) Superior temporal sulcus; (f) Inferior temporal sulcus; (g) Intra parietal sulcus; (h) Transverse occipital sulcus rest of the hemisphere. On the LHs, the Sylvian fissure is centralised superior temporal sulcus on the LHs showed higher variation rates (red) rostrally compared to the longitudinal distribution on the RHs (blue) towards the parietal lobe (Figure 3e). Similarly, the inferior tem- (Figure 3d). The anterior horizontal ramus and the ascending ramus of poral sulcus could be seen more centralised (red) to the temporal the Sylvian fissure can be seen as centralised (red) distributions on the poles with higher variation (blue) towards the occipital lobe on both inferior extremities of the frontal lobe. More extreme variations (blue) hemispheres (Figure 3f). Furthermore, the intraparietal sulcus showed of these rami could be seen on the LHs compared to the RHs two clusters of data; however, the most common area of identification (Figure 3a,b). was towards the caudal extremities of the parietal lobe (Figure 3g). On the temporal lobes, the superior temporal sulcus could be Finally, as for the occipital sulci from most rostral to caudal, the seen directly inferior to the impression of the Sylvian fissure showing transverse occipital sulcus showed a centralised distribution (red) on a more centralised (red) distribution towards the temporal pole. The the superior lateral extremities of the occipital lobe with more de JAGER ET AL. 4439 posterior extremities of the occipital pole, with variation in the lateral calcarine sulcus stretching laterally on the occipital pole as well as extreme variation seen in the position retro-calcarine sulcus (Figure 4c,d). The distributions of the major sulci of the frontal lobe overlap in the RHs and in the LHs (Figure 5a). Additionally, overlapping can be observed when viewing the precentral-, central- and postcentral sulci, particularly towards the superior margin on the LHs (Figure 5b). Sig- nificant overlapping was observed between the density maps of the anterior horizontal ramus and the ascending ramus of the Sylvian fis- sure (Figure 5c), and, on the occipital pole, between the distributions of the lunate sulcus, the inferior occipital sulcus, the retro-calcarine sulcus and the lateral calcarine sulcus (Figure 5d). 4 | DI SCU SSION This study provides the first quantitative description of sulcal imprint variation patterns on extant human endocasts. Our method used a density-based approach combined with a user-independent sulcal imprint detection method, implemented on 20 samples. As indicated by the red areas on the density maps, sulcal imprints of the frontal and temporal lobes, and the Sylvian fissure, respectively, could be identified in regions where they are usually found in the brain (Connolly, 1950; Ono et al., 1990). Interestingly, the middle frontal sulcus was most frequently identified as having a rostral- and caudal segment, as described by Connolly (1950) in human brains, and more recently reiterated by Petrides and Pandya (2004) on brain MRIs. The fronto-marginal sulcus was mainly recognised as a horizontally orien- tated impression, extending laterally from the midline. The fronto- orbital sulcus imprint was observed on the endocasts just anterior to “Broca's cap”, despite the identification of this sulcus in humans being a rather controversial topic (Connolly, 1950; Falk et al., 2018; Ono et al., 1990; addressed in de Jager et al., 2019). It is evident from the density maps that higher variation can be observed in the anterior horizontal ramus and the ascending ramus of the Sylvian fissure on the LHs, which correlated with findings of asymmetrical variability by Fillard et al. (2007). High variability in density patterns observed for the precentral, central and postcentral sulcal impressions could result from distortion F I G U R E 4 Density map indicating the distribution of sulcal caused on the endocast by cranial impressions of the middle menin- imprints around the mean in left and right views. Red indicates the geal vessels, causing missing information (Bruner et al., 2018). Despite mean distribution and blue indicates mean + 2 standard deviations. the distortion experienced, it is evident from the density maps that (a) Lunate sulcus; (b) Inferior occipital sulcus; (c) Lateral calcarine sulcus; (d) Retro-calcarine sulcus the inferior part of the precentral sulcus was reliably identified in most cases, represented by the mean observation in red. The detail with which the regional imprints were detected by our imaging modalities variation being visible rostral on the LHs (Figure 3h). The lunate sulcus contributed towards our ability to identify the sulcal imprint of inter- is centralised (red) in the middle of the occipital lobe directly caudal to est, thereby reducing the impact of this potential limitation. Similarly, the impression of the lambdoid suture (Figure 4a). The lateral/inferior we could not reliably determine the postcentral sulcus, which may be occipital sulci showed an obvious distribution close to the inferior- due to the complex branching networks of blood vessels covering this lateral border of the occipital lobes and shows higher variation on the area, and as a result, this feature showed highly variable distribution RHs (Figure 4b). Furthermore, the lateral calcarine sulcus and the on the density map. High surface variability of the parietal cortex, retro-calcarine sulcus were identified showing distributions on the housing the postcentral sulcus, was reported on MRI studies (Fillard 4440 de JAGER ET AL. F I G U R E 5 Overlap of density maps. (a) Frontal lobe showing the superior frontal sulcus (yellow), middle frontal sulcus (pink) and fronto- marginal sulcus (red); (b) Left and right parietal lobes showing precentral sulcus (green), central sulcus (yellow) and postcentral sulcus (orange); (c) Left and right Broca's area showing anterior horizontal ramus (orange) and ascending ramus (green); (d) Occipital lobe showing the lunate sulcus (red), inferior occipital sulcus (pink), lateral calcarine sulcus (yellow) and retro-calcarine sulcus (green)  mez-Robles et al., 2013). Cerebrospinal fluid (CSF) and et al., 2007; Go substantial cortical expansion. In particular, because of the link blood vessels usually fill the main sulci that delineate functional areas, between the Broca's cap and language, the presence or absence of which, may, subsequently, not be well reproduced on endocasts the primitive fronto-orbital sulcus of the human-like horizontal and  n, 2013). Although we do not know the handed- (Zollikofer & De Leo ascending rami has been the source of essential discussions. Because ness of our sample, it is nonetheless clear that the central sulcus on of the presence of the fronto-orbital sulcus in early hominins (Carlson the LHs was more distinctly identified than on RHs, which could be et al., 2011; Holloway, Broadfield, & Yuan, 2004), it has long been due to the central sulcus more prominently imprinting on the inner thought that the human-like pattern emerged along with the earliest surface of the braincase in right-hand dominant individuals, based on humans (rev. in Beaudet, 2021). However, a recent revision of the findings from Amunts et al. (2000). More research is required to human fossil record rejected this hypothesis, instead suggesting a explore this hypothesis. later emergence of the modern human configuration of the Broca's Based on the distribution seen on the density model, the transverse  n et al., 2021). Interestingly, our study demonstrates cap (Ponce de Leo occipital sulcus can easily be misidentified on the endocasts due to the that the distribution of these critical sulci in the Broca's region overlap tendency of the intraparietal sulcus to terminate posteriorly, close to the significantly (Figure 5). As such, we might be cautious when describing lambdoid suture (refer to Figure 3). This suture frequently creates phan- this region in the fossil record. The frontal sulci are often described in tom markings on the cranial endocast of extant humans and proves help- the fossil record due to the good quality of imprints found in this ful in indicating the position of the superior extremities of the parieto- region of the endocast (e.g. Falk, 2014). The presence and configura- occipital sulcus at the angle formed by the sagittal suture and the lam- tion of the middle sulci have been used to identify more derived pat- bdoid suture, a landmark often used in neurosurgery (Ribas et al., 2006). terns in the endocasts of early hominins (Falk, 2014). Our maps Asymmetry was noted when observing the density maps of the lateral demonstrate that the positions of the superior and middle sulci are calcarine sulcus. This may be due to distortion (created) by the right dom- relatively consistent. However, a recent re-analysis of the chimpanzee inant dural venous sinus groove (García-Tabernero et al., 2018). As brain questioned the use of these sulci in evolutionary discussions addressed by de Jager et al. (2019), age was excluded as a variable in this (Falk et al., 2018). One of the “hottest” debates concerning the identi- study due to sample size constraints. Based on Liu et al. (2010) proposal, fication of sulci in fossil hominins is probably the one surrounding the no new brain expansion occurs in older individuals. lunate sulcus. Using fossil specimens from South Africa, scientists Some of these sulci have been intensely discussed in the fossil have been arguing about a more derived or more ancestral position of record because of their functional role or use as a proxy to indicate this structure for decades (Falk, 1983, 2014; Holloway, 1981; de JAGER ET AL. 4441 Holloway, Clarke, & Tobias, 2004). Finding the correct position of the which this method afforded imprint detection will significantly con- lunate sulcus is of prime interest, since a more posterior (derived) tribute to future studies as it enables the user to virtually interact with position might indicate a significant expansion of the parietal associa- a comparative map of extant human sulcal imprints. Furthermore, our tion cortex. Our results clearly show a substantial overlap of the model correlates well with sulcal variation previous studies on brains occipital sulci, including the lunate sulcus, which unequivocally accen- and brain-MRIs, in particular the sulci of the frontal lobes and occipital tuates the risk of mis-identifying this structure in the fossil record. lobes. This map will consequently prove anatomically invaluable for Expanding the atlas to include more samples will likely only improve future studies (and even revisiting previous studies) pertaining to the the robustness of our density maps, which already compare well with fossil record. Nevertheless, it is essential to note that cranial endo- direct brain studies (Fillard et al., 2007; Iaria & Petrides, 2007; Juch casts present with significant expanses of missing cortical sulcal infor- et al., 2005; Ochiai et al., 2004). mation. This is particularly true when studying cortical imprints on Recent MRI studies on the variability of brain sulci suggest higher extant human endocasts, due to the lack of imprint formation in cer- variation and asymmetry in sulcal patterns associated with high order tain areas which is caused surrounding anatomical structures association areas (e.g. prefrontal cortex and parietal cortex) and lower (e.g. meninges). Furthermore, the sequence of emergence of second- variability in sulci delimiting the cerebral lobes such as the Sylvian fis- ary and tertiary sulci during development of the brain could have  mez-Robles sure and the central sulcus (Fillard et al., 2007; Go effect on the clarity of imprints formed. Fossil samples present the et al., 2013, 2015). The findings from these authors correlate with our added complication of sulcal imprint detection occasionally being density map data which show higher variability in imprints of the mid- compromised by the poor preservation of these features during dle frontal sulcus, precentral sulcus, postcentral sulcus, intraparietal fossilisation. Future studies will address this “obstacle” of missing sulcus and sulci of the temporal lobe especially the superior- and infe- information of cortical imprints, using advanced statistical methods to rior temporal sulcus on the parietal lobe. In contrary to the above- develop a model that can be used to contribute to the statistical mentioned studies, our results show less variation on imprints of the robusticity of our atlas of sulcal imprints. Sylvian fissure compared to the imprints of the central sulcus, and our results do not show the unilateral extension of the Sylvian fissure ACKNOWLEDG MENTS  mez-Robles et al. onto the LH's parietal lobe such as the findings of Go We thank G. Krüger for collection access; AN van Schoor for anatomi- (2013). Specifically considering the prefrontal cortex, Juch et al. cal expertise; F. de Beer and L. Bam (Necsa) for data acquisition; (2005) reported the origin of superior frontal sulcus to connect to J. Dumoncel (AMIS – UMR5288) for technical support. Supported by precentral sulcus in at least 74% of their sample, which does not show the South Africa/France (PROTEA) Joint Research Programme (grant in our sample due to the lack of imprint formation of the superior part number 129923) and by the University of Cambridge Harding Distin- of precentral sulcus. Our data does however show possible overlap guished Postgraduate Scholars Programme; we thank the DST-NRF between inferior frontal-, precentral- and fronto-orbital sulci which for sponsoring the Micro-XCT facility at Necsa. Ethical clearance was correlates with the variation seen in previous studies on brains (Juch obtained from the Main Research Ethics committee of the Faculty of et al., 2005; Ono et al., 1990). Juch et al. (2005) identified frequent Health Science, University of Pretoria August 2017 (No: 339/2017). juxtaposition of the ascending ramus and precentral sulcus, which cor- Additional ethical clearance was obtained from the ethical committee relates with our findings on extant human endocasts. The superior at Cambridge. temporal sulcus is known for its segmental patterns (Ochiai et al., 2004), in our findings the imprint of the superior temporal sul- CONFLIC T OF INT ER E ST cus does not indicate the segmented patterns but instead one promi- The authors have no conflict of interest to declare. nent imprint distribution starting at the temporal pole and becoming more variable towards the parietal cortex. The opposite can be DATA AVAILABILITY STAT EMEN T observed with the imprint of the inferior temporal sulcus which indi- The data that support the findings of this study are available from the cates more than one clustering of imprints, correlating with descrip- corresponding author upon reasonable request. tions by Ono et al. (1990), with the most prominent imprint distribution being closer to the temporal pole. Finally, the variation of OR CID sulcal imprints on the occipital lobe from our data correlates well with Edwin John de Jager https://orcid.org/0000-0003-3199-8566 probability maps from brain-MRIs studied by Iaria and Petrides (2007), Caroline Fonta https://orcid.org/0000-0001-6326-4044 which similarly reported significant blending between the lunate, the Amélie Beaudet https://orcid.org/0000-0002-9363-5966 transverse occipital and the lateral occipital sulcus. Similarities of observations made between brains and our results on endocasts sup- RE FE RE NCE S port the suggestion that endocasts remain a good proxy for studying Amunts, K., Jäncke, L., Mohlberg, H., Steinmetz, H., & Zilles, K. (2000). sulcal imprint variation. Interhemispheric asymmetry of the human motor cortex related to handedness and gender. Neuropsychologia, 38(3), 304–312. https:// Our methodological approach provides the first anatomically doi.org/10.1016/S0028-3932(99)00075-5 derived map for sulcal imprint variation on extant human endocasts Beaudet, A. (2021). The enigmatic origins of the human brain. Science, using an automated sulcal imprint detection method. The detail with 372(6538), 124–125. https://doi.org/10.1126/science.abi4661 4442 de JAGER ET AL. Beaudet, A., Du, A., & Wood, B. (2019). Evolution of the modern human Fiot, J.-B., Raguet, H., Risser, L., Cohen, L. D., Fripp, J., & Vialard, F. X. brain. In M. A. Hofman (Ed.), Progress in brain research (250th ed., (2014). Longitudinal deformation models, spatial regularizations and pp. 219–250). Elsevier B.V. https://doi.org/10.1016/bs.pbr.2019. learning strategies to quantify Alzheimer's disease progression. 01.004 NeuroImage: Clinical, 4, 718–729. https://doi.org/10.1016/j.nicl.2014. Beaudet, A., Dumoncel, J., de Beer, F., Durrleman, S., Gilissen, E., 02.002 Hoffman, J., Oettlé, A., Subsol, G., Thackeray, J. F., & Braga, J. (2016). García-Tabernero, A., Peña-Melián, A., & Rosas, A. (2018). Primary visual Morphoarchitectural variation in the extant human endocast. Journal cortex in neandertals as revealed from the occipital remains from the of Human Evolution, 101, 65–78. El Sidro n site, with emphasis on the new SD-2300 specimen. Journal Beaudet, A., & Gilissen, E. (2018). Fossil primate Endocasts: Perspectives of Anatomy, 233(1), 33–45. https://doi.org/10.1111/joa.12812 from advanced imaging techniques. In E. Bruner, N. Olgihara, & H. Glasser, M. F., Coalson, T. S., Robinson, E. C., Hacker, C. D., Harwell, J., Tanebe (Eds.), Digital endocasts: From skulls to brains (pp. 47–58). Yacoub, E., Ugurbil, K., Andersson, J., Beckmann, C. F., Jenkinson, M., Springer. https://doi.org/10.1007/978-4-431-56582-6_4 Smith, S. M., & Van Essen, D. C. (2016). A multi-modal parcellation of Bookstein, F. L. (1989). Principal warps: Thin-plate splines and the decomposi- human cerebral cortex. Nature, 536(7615), 171–178. https://doi.org/ tion of deformations. IEEE Transactions on Pattern Analysis and Machine 10.1038/nature18933 Intelligence, 11(6), 567–585. https://doi.org/10.1109/34.24792 Go mez-Robles, A., Hopkins, W. D., Schapiro, S. J., & Sherwood, C. C. Bruner, E., Ogihara, N., & Tanabe, H. C. (2018). In E. Bruner, N. Ogihara, & (2015). Relaxed genetic control of cortical organization in human H. C. Tanabe (Eds.), Digital endocasts: From skulls to brains (1st ed.). brains compared with chimpanzees. Proceedings of the National Acad- Springer Japan. https://doi.org/10.1007/978-4-431-56582-6 emy of Sciences of the United States of America, 112(48), 14799– Carlson, K. J., Stout, D., Jashashvili, T., de Ruiter, D. J., Tafforeau, P., 14804. https://doi.org/10.1073/pnas.1512646112 Carlson, K., & Berger, L. R. (2011). The Endocast of MH1, Go mez-Robles, A., Hopkins, W. D., & Sherwood, C. C. (2013). Increased Australopithecus sediba. Science, 333(6048), 1402–1407. https://doi. morphological asymmetry, evolvability and plasticity in human brain org/10.1126/science.1203922 evolution. Proceedings of the Royal Society B-Biological Sciences, Connolly, C. J. (1950). External morphology of the primate brain. Charles 280(1761). https://doi.org/10.1098/rspb.2013.0575 C. Thomas. Holloway, R. L. (1978). The relevance of Endocasts for studying primate de Jager, E. J., van Schoor, A. N., Hoffman, J. W., Oettlé, A. C., Fonta, C., brain evolution. In C. R. Noback (Ed.), Sensory systems of primates Mescam, M., Risser, L., & Beaudet, A. (2019). Sulcal pattern variation (pp. 181–200). Springer. https://doi.org/10.1007/978-1-4684-2484- in extant human endocasts. Journal of Anatomy, 235(4), 803–810. 3_8 https://doi.org/10.1111/joa.13030 Holloway, R. L. (1981). Revisiting the south African Taung australopithe- Dumoncel, J., Subsol, G., Durrleman, S., Bertrand, A., Jager, E., cine endocast: The position of the lunate sulcus as determined by the Oettlé, A. C., Lockhat, Z., Suleman, F. E., & Beaudet, A. (2021). Are stereoplotting technique. American Journal of Physical Anthropology, endocasts reliable proxies for brains? A 3D quantitative comparison of 56(1), 43–58. https://doi.org/10.1002/ajpa.1330560105 the extant human brain and endocast. Journal of Anatomy, 238(2), Holloway, R. L. (2009). Brain fossils: Endocasts. Encyclopedia of Neurosci- 480–488. https://doi.org/10.1111/joa.13318 ence, 2009, 353–361. https://doi.org/10.1016/B978-008045046-9. Duvernoy, H. M., & Bourgouin, P. (1999). The human brain: Surface, three- 00941-4 dimensional sectional anatomy with MRI, and blood supply. Springer Holloway, R. L., Broadfield, D. C., & Yuan, M. S. (2004). In J. H. Schwartz & Wien. I. Tattersall (Eds.), The human fossil record (Vol. 3). John Wiley & Sons, Evans, A. C., Collins, D. L., Neelin, P., MacDonald, D., Kamber, M., & Inc. https://doi.org/10.1002/0471663573 Marett, T. S. (1994). Three-dimensional correlative imaging: Applica- Holloway, R. L., Clarke, R. J., & Tobias, P. V. (2004). Posterior lunate sulcus tions in human brain mapping. In R. Thatcher, M. Hallet, T. Zeffiro, E. in Australopithecus africanus: Was dart right? Comptes Rendus Palevol, John, & M. Huerta (Eds.), Functional neuroimaging: Technical founda- 3(4), 287–293. https://doi.org/10.1016/j.crpv.2003.09.030 tions (pp. 145–162). Academic Press. Iaria, G., & Petrides, M. (2007). Occipital sulci of the human brain: Variabil- Falk, D. (1980a). Comparative study of the Endocranial casts of new and ity and probability maps. The Journal of Comparative Neurology, 501(2), Old World monkeys. In R. L. Ciochon & A. B. Chiarelli (Eds.), Evolution- 243–259. https://doi.org/10.1002/cne.21254 ary biology of the New World monkeys and continental drift (pp. 275– Juch, H., Zimine, I., Seghier, M. L., Lazeyras, F., & Fasel, J. H. D. (2005). 292). Springer. https://doi.org/10.1007/978-1-4684-3764-5_13 Anatomical variability of the lateral frontal lobe surface: Implication for Falk, D. (1980b). A reanalysis of the south African australopithecine natu- intersubject variability in language neuroimaging. NeuroImage, 24(2), ral endocasts. American Journal of Physical Anthropology, 53(4), 525– 504–514. https://doi.org/10.1016/j.neuroimage.2004.08.037 539. https://doi.org/10.1002/ajpa.1330530409 Karcher, H. (1977). Riemannian center of mass and mollifier smoothing. Falk, D. (1983). Cerebral cortices of east African early hominids. Science, Communications on Pure and Applied Mathematics, 30(5), 509–541. 221(4615), 1072–1074. L'Abbé, E. N., Loots, M., & Meiring, J. H. (2005). The Pretoria bone collec- Falk, D. (1987). Hominid paleoneurology. Annual Review of Anthropology, tion: A modern south African skeletal sample. Homo, 56(2), 197–205. 16, 13–30. https://doi.org/10.1016/j.jchb.2004.10.004 Falk, D. (2014). Interpreting sulci on hominin endocasts: Old hypotheses Le Gros Clark, W. E., Cooper, D. M., & Zuckerman, S. (1936). The and new findings. Frontiers in Human Neuroscience, 8, 1–11. https:// Endocranial cast of the chimpanzee. The Journal of the Royal Anthropo- doi.org/10.3389/fnhum.2014.00134 logical Institute of Great Britain and Ireland, 66, 249–268. https://doi. Falk, D., Zollikofer, C. P. E., Ponce De Leo n, M., Semendeferi, K., Alatorre org/10.2307/2844081 Warren, J. L., & Hopkins, W. D. (2018). Identification of in vivo sulci on Liu, T., Wen, W., Zhu, W., Trollor, J., Reppermund, S., Crawford, J., the external surface of eight adult chimpanzee brains: Implications for Jin, J. S., Luo, S., Brodaty, H., & Sachdev, P. (2010). The effects of age interpreting early hominin Endocasts. Brain, Behavior and Evolution, and sex on cortical sulci in the elderly. NeuroImage, 51(1), 19–27. 91(1), 45–58. https://doi.org/10.1159/000487248 https://doi.org/10.1016/j.neuroimage.2010.02.016 Fillard, P., Arsigny, V., Pennec, X., Hayashi, K. M., Thompson, P. M., & Lockwood, C. A., & Tobias, P. V. (1999). A large male hominin cranium Ayache, N. (2007). Measuring brain variability by extrapolating sparse from Sterkfontein, South Africa, and the status of Australopithecus tensor fields measured on sulcal lines. NeuroImage, 34(2), 639–650. africanus. Journal of Human Evolution, 36(6), 637–685. https://doi.org/ https://doi.org/10.1016/j.neuroimage.2006.09.027 10.1006/jhev.1999.0299 de JAGER ET AL. 4443 Moss, M. L., & Young, R. W. (1960). A functional approach to craniology. Thompson, P. M., Woods, R. P., Mega, M. S., & Toga, A. W. (2000). American Journal of Physical Anthropology, 18(4), 281–292. https://doi. Mathematical/computational challenges in creating deformable and org/10.1002/ajpa.1330180406 probabilistic atlases of the human brain. Human Brain Mapping, 9(2), Neubauer, S. (2014). Endocasts: Possibilities and limitations for the inter- 81–92. https://doi.org/10.1002/(SICI)1097-0193(200002)9:2<81:: pretation of human brain evolution. Brain, Behavior and Evolution, AID-HBM3>3.0.CO;2-8 84(2), 117–134. https://doi.org/10.1159/000365276 Toga, A. W., & Thompson, P. M. (2001). Maps of the brain. The Anatomical Ochiai, T., Grimault, S., Scavarda, D., Roch, G., Hori, T., Rivière, D., Record, 265(2), 37–53. https://doi.org/10.1002/ar.1057 Mangin, J. F., & Régis, J. (2004). Sulcal pattern and morphology of the Toga, A. W., Thompson, P. M., Mori, S., Amunts, K., & Zilles, K. (2006). superior temporal sulcus. NeuroImage, 22(2), 706–719. https://doi. Towards multimodal atlases of the human brain. Nature Reviews Neuro- org/10.1016/j.neuroimage.2004.01.023 science, 7(12), 952–966. https://doi.org/10.1038/nrn2012 Ono, H., Kubik, S., & Abernathey, C. (1990). Atlas of cerebral sulci. Vialard, F.-X., Risser, L., Rueckert, D., & Cotter, C. J. (2012). Diffeomorphic G. Thieme Verlag. 3D image registration via geodesic shooting using an efficient Adjoint Petrides, M., & Pandya, D. N. (2004). The Frontal Cortex. In J. Mai & calculation. International Journal of Computer Vision, 97(2), 229–241. G. Paxinos (Eds.), The Human Nervous System (3rd edn., pp. 950–972). https://doi.org/10.1007/s11263-011-0481-8 Elsevier. https://doi.org/10.1016/B978-012547626-3/50026-0 Vialard, F.-X., Risser, L., Rueckert, D., & Holm, D. D. (2012). Diffeomorphic Ponce de Leo  n, M. S., Bienvenu, T., Marom, A., Engel, S., Tafforeau, P., atlas estimation using geodesic shooting on volumetric images. Annals Alatorre Warren, J. L., Lordkipanidze, D., Kurniawan, I., Murti, D. B., of the British Machine Vision Association, 2012(5), 1–12. Suriyanto, R. A., Koesbardiati, T., & Zollikofer, C. P. E. (2021). The Yoshizawa, S., Belyaev, A., Yokota, H., & Seidel, H. P. (2007). Fast and primitive brain of early homo. Science, 372(6538), 165–171. https:// faithful geometric algorithm for detecting crest lines on meshes. 15th doi.org/10.1126/science.aaz0032 Pacific Conference on Computer Graphics and Applications (PG'07), 2, Ribas, G. C. (2010). The cerebral sulci and gyri. Neurosurgical Focus, 28(2), 231–237. https://doi.org/10.1109/PG.2007.24 E2. https://doi.org/10.3171/2009.11.FOCUS09245 Yoshizawa, S., Belyaev, A., Yokota, H., & Seidel, H. P. (2008). Fast, robust, Ribas, G. C., Yasuda, A., Ribas, E. C., Nishikuni, K., & Rodrigues, A. J. and faithful methods for detecting crest lines on meshes. Computer (2006). Surgical anatomy of microneurosurgical Sulcal key points. Aided Geometric Design, 25(8), 545–560. https://doi.org/10.1016/j. Operative Neurosurgery, 59(4), 177–210. https://doi.org/10.1227/01. cagd.2008.06.008 NEU.0000240682.28616.b2 Zollikofer, C. P. E., & De Leo  n, M. S. P. (2013). Pandora's growing box: Rivière, D., Geffroy, D., Denghien, I., Souedet, N., & Cointepas, Y. (2009). Inferring the evolution and development of hominin brains from endo- BrainVISA: An extensible software environment for sharing multi- casts. Evolutionary Anthropology: Issues, News, and Reviews, 22(1), 20– modal neuroimaging data and processing tools. NeuroImage, 47(1), 33. https://doi.org/10.1002/evan.21333 S163. https://doi.org/10.1016/S1053-8119 Stiles, J., & Jernigan, T. L. (2010). The basics of brain development. Neuro- psychology Review, 20(4), 327–348. https://doi.org/10.1007/s11065- 010-9148-4 SUPPORTING INF ORMATION Subsol, G., Gesquière, G., Braga, J., & Thackeray, F. (2010). 3D automatic Additional supporting information may be found in the online version methods to segment virtual endocasts: State of the art and future of the article at the publisher's website. directions. American Journal of Physical Anthropology, 141, 226–227. Thompson, P., & Toga, A. W. (1999). Anatomically driven strategies for high-dimensional brain image warping and pathology detection. In W. How to cite this article: de Jager, E. J., Risser, L., Mescam, M., Toga (Ed.), Brain warping (pp. 311–336). Elsevier. https://doi.org/10. 1016/B978-012692535-7/50094-X Fonta, C., & Beaudet, A. (2022). Sulci 3D mapping from human Thompson, P. M., & Toga, A. W. (1997). Detection, visualization, and ani- cranial endocasts: A powerful tool to study hominin brain mation of abnormal anatomic structure with a deformable probabilistic evolution. Human Brain Mapping, 43(14), 4433–4443. https:// brain atlas based on random vector field transformations. Medical doi.org/10.1002/hbm.25964 Image Analysis, 1(4), 271–294. https://doi.org/10.1016/S1361-8415 (97)85002-5