Cross-sectional anatomy, computed tomography, and magnetic resonance imaging of the banded houndshark (Triakis scyllium)

Due to their important phylogenetic position among extant vertebrates, sharks are an invaluable group in evolutionary developmental biology studies. A thorough understanding of shark anatomy is essential to facilitate these studies and documentation of this iconic taxon. With the increasing availability of cross-sectional imaging techniques, the complicated anatomy of both cartilaginous and soft tissues can be analyzed non-invasively, quickly, and accurately. The aim of this study is to provide a detailed anatomical description of the normal banded houndshark (Triakis scyllium) using computed tomography (CT) and magnetic resonance imaging (MRI) along with cryosection images. Three banded houndsharks were scanned using a 64-detector row spiral CT scanner and a 3 T MRI scanner. All images were digitally stored and assessed using open-source Digital Imaging and Communications in Medicine viewer software in the transverse, sagittal, and dorsal dimensions. The banded houndshark cadavers were then cryosectioned at approximately 1-cm intervals. Corresponding transverse cryosection images were chosen to identify the best anatomical correlations for transverse CT and MRI images. The resulting images provided excellent detail of the major anatomical structures of the banded houndshark. The illustrations in the present study could be considered as a useful reference for interpretation of normal and pathological imaging studies of sharks.

In the same context, precise atlases of CT and MRI findings in shark species should also be established. The banded houndshark (Triakis scyllium) is a relatively small-sized shark that is easily found in Korean water. They inhabit the Northwest Pacific Ocean and are classified as the "least concern" group by the International Union for Conservation of Nature (IUCN) red list 4,39 . The purpose of this study is to set up a detailed atlas of the banded houndshark by comparing CT and MRI images with actual cryosections. To the best of our knowledge, this is the first study reporting delicate comparison of CT, MRI, and cryosection findings for shark anatomy.

Discussion
The MRI signal intensity is proportional to the hydrogen density of the structures, ensuring excellent definition of soft tissues, organs, and cavitary structures in banded houndsharks. These structures included major muscles; cardiovascular system structures, including the atrium, ventricle, and major arteries and veins; digestive system structures, including the stomach, gall bladder, and valvular intestine; excretory system structures, including the kidney; nervous system structures, including the cerebrum, optic lobe, and thalamus. Due to the absence of hydrogen atoms, most of the cartilaginous skeletons in banded houndsharks did not provide sufficient magnetic resonance signal and showed a hypointense (black/dark grey) pattern. Nonetheless, some cartilaginous structures, such as the jaws and coracoid bar, could still be observed due to the contrast between the cartilage and the adjacent soft tissues. In contrast to MRI, excellent discrimination of cartilaginous structures was evident in the CT images. The higher electron density of cartilages caused increased attenuation compared to soft tissue, making them appear more whitish in the CT images. The cartilage margins were better defined by means of CT than cryosections or MRI, particularly in assessing joint structures. Soft tissues were vaguely distinguished using CT examinations due to their similar attenuation properties. Nevertheless, attenuation differences between lipid-rich tissues (e.g., liver) and muscles (e.g., stomach and valvular intestine) were shown in varying shades of grey allowing definition of some organs in the visceral cavity.   www.nature.com/scientificreports/ CT and MRI offer considerable advantages over traditional radiographic approaches on identification of the anatomical structures of sharks. In traditional radiography, images are presented as superimposed two-dimensional projections of three-dimensional structures, which result in limitation to exhibit high conspicuity. The tomographic nature of CT and MRI allows organs to be examined in thin sections, eliminating superimposition of overlying structures that may hinder the specific interpretation of anatomical structures.
Another major advantage of CT and MRI is the ability to reformat the datasets in any imaging plane or as 3D projections, allowing better understanding of the spatial relationships of anatomical structures in sharks. These techniques have added detail of the anatomical structures and can be used as an anatomic reference for imaging studies of the banded houndshark and other shark species. It is recommended to reduce CT/MRI slice thickness as much as possible in order to achieve excellent quality and high detail of 3D reconstructed images, since increasing slice thicknesses may require more interpolation between slices when rendering 3D images, which results in a loss of resolution 40 .
Due to their important phylogenetic position among extant vertebrates, sharks are an invaluable group in evolutionary developmental biology studies. However, the presence of high levels of urea accelerates the decomposition of shark flesh 41 . Moreover, because of its cartilaginous nature, shark skeleton has a high tendency to warp, crack, and shrink during specimen preparation 42 . These factors represent major difficulties in preserving anatomical data from sharks. In the present study, CT and MRI were used to identify and document soft tissues and cartilaginous structures in banded houndsharks. The non-invasive nature of CT and MRI prevented the irretrievable destruction of specimens during traditional dissection. The spatial relationships among organs, soft tissues, and cartilage can be selectively observed in situ in their natural locations 43,44 . After examinations, all images are permanently recorded in the Digital Imaging and Communications in Medicine (DICOM) format and could be recalled at will. The images are digitally transferable, which can also facilitate discussion and opinion sharing among professionals worldwide even if the specimens cannot be physically provided.
Banded houndshark is a relatively small-sized shark species that can be adequately fit into the bore of medical CT and MRI used in the present study. However, to the best of the authors' knowledge, only the head of large-sized shark species has been subjected to imaging examination possibly because of limited gantry size [45][46][47] . Sharks with body girth larger than the maximum diameter of field of view (FOV) of medical CT and MRI may  www.nature.com/scientificreports/ also induce out-of-field artifacts, anatomical structures that lies beyond the FOV may be truncated. With the rapid advancement of medical imaging technologies, wide bore CT and MRI have become available for large animals [48][49][50] . Although the current availability of wide bore CT and MRI scanners are limited and the set up cost is high, the use of wide bore CT and MRI may offer a feasible alternative for imaging examination of large-sized shark species in future.

Conclusions
The illustrations in the present study are the first to provide a comprehensive atlas, using CT and MRI, to evaluate the anatomical structures of the banded houndshark. Both imaging modalities provide good contrast for the anatomical structures of the banded houndshark. With the rapid advancement and increasing availability of medical imaging technologies, CT and MRI are enhancing our knowledge of the anatomy of sharks together         Ages of the sharks were estimated based on their TBL and it was judged that the sharks were about to be sexually mature 51 . All animals were selected by a veterinarian (SWK) based on the lack of external evidence of trauma and disease. The banded houndsharks were euthanized via tricaine methanesulfonate overdose (500 ppm) right before imaging examinations in order to minimize post-mortem changes.
Magnetic resonance imaging (MRI). MRI examinations were performed using a 3-T (T) MRI scanner (Achieva, Philips Healthcare, The Netherlands). A head coil was used to receive the signal using fast spin echo sequences in the T2-weighted (T2W) mode, in which signals from loosely bound protons (e.g., fluids) were enhanced: www.nature.com/scientificreports/ Cryosectioning. All specimens were frozen (− 22 °C) in the ventral recumbency position until cryosectioning could be carried out. The banded houndsharks were sectioned using an electrical band saw along the transverse plane at approximately 1-cm intervals. Slices were numbered, cleaned, and photographed on both sides. Fourteen transverse section levels were selected. For each section level, the corresponding transverse MRI and CT images were chosen to identify the best anatomical correlation. Lines corresponding to these levels were superimposed on three-dimensional reconstructed images of an banded houndshark (Fig. 25).
Three-dimensional (3D) reconstruction. 3D reconstructions of the skeletons were made on acquired CT data sets using intensity-based segmentation methods of the inbuilt software of 64-detector row spiral CT scanner (Aquilion, Toshiba Medical Systems, Nasu, Japan). The reconstructed skeletons were visualised after excluding remaining residual soft tissue structures. 3D reconstructions of major organs were made on MRI data sets using intensity-based and manual segmentation methods of an open-source DICOM viewer-InVesalius 3.0 (CTI-Center for Information and Technology Renato Archer, Campinas, São Paulo, Brazil).
Ethical approval. The study was approved by the Seoul National University Institutional Animal Care and Use Committee (approval number: SNU-190925-2). All experiments were performed in accordance with the IACUC guidelines and regulations. This study was also performed in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines: to reflect the scarcity of banded houndshark over 90 cm TBL in the fish markets of the Republic of Korea, sample size was decided to three; to minimize the potential confounders, both male and female sharks were used; to show the developmental stages, TBL and estimated age of each shark were provided.