Ct atlas

Author: s | 2025-04-24

CT-based Atlas of the Head and Neck The head and neck atlas was derived from a reduced resolution (256x256) CT MANIXdata set from the OSIRIX data sets. New Atlas Viewer. CT

CT ATLAS Training for NCCs

With MRI, a precise, individual segmentation is time consuming and not feasible in clinical routines.A possible solution to obtain masks of anatomical structures is to use an already existing brain atlas. In atlas-based approaches, a template intensity image is registered to the target image, and the resulting deformation is then applied to the anatomical labels of the template to match the target space. In this way, the existing information of the atlas image can be transferred to the image of interest [1]. To better address the variability between scans of different subjects, multi-atlas registration can be used [2]. With this approach, the target image is registered to multiple atlas images and the resulting label images are combined via majority voting algorithms [2,3,4].There are multiple MR-based anatomical atlases available to transfer different labeled regions to new unlabeled images. To the best of our knowledge, there is no detailed CT-based anatomical brain atlas, and few registration-based approaches have been proposed for transferring anatomical information from an MR image to a CT scan. This is a challenging task and requires reliable multi-modal, inter-subject, non-rigid CT-MR registration. In the approaches proposed by [5,6,7], the task is reduced to a mono-modal registration problem by decomposing it into two steps. First, CT images are synthesized into MRI images, and then the registration is performed between the synthesized MRI images and the MRI atlas.Moreover, in [8] the authors propose a method to create an average CT atlas. Therefore, they first create an average CT image and then register multiple MRI atlas images to the average image. Finally, the anatomical labels from the MRI atlases are fused completing the final average CT atlas.Furthermore, the authors in [9] investigate an atlas-based method to segment ventricles in CT images. However, when it comes to CT images, it would be beneficial to delineate a larger number of anatomical structures. The authors in [10] propose a registration-based method to build a CT head atlas with anatomical structures for the Chinese population. However, they manually correct the segmentation and do not use MR atlas images. Direct multi-modal registration usually is eased by utilizing additional information and correspondence structures in the distance measure computation. Gao et al. [11] extract the midsagittal plane and use brain surface matching. Chen et al. [12] propose a combination of landmarks and mutual information (MI) as similarity measure to include local and global anatomical structure. Learning-based approaches, as described

CT Angio Atlas - neuroangio.org

In [4], aim to overcome the disadvantages of multi-modal similarity measures such as MI. To evaluate the performance of the label propagation on MRI scans, Dubost et al. [13] introduce the computation of the overlap from automatically segmented ventricles and choose the result with the highest Dice score.In this work, we present a novel enhanced multi-modal, multi-atlas registration approach to propagate anatomical labels from an MRI atlas to new unseen CT brain scans. Our approach builds on deformable registration utilizing corresponding structures (brain parenchyma and ventricles) as extra input to guide the registration process between the CT image and the MR atlas. On that account, we propose a convolutional neural network solution to automatically segment brain volume and ventricle system in CT images.Methods and materialOur solution is divided into three steps. Firstly, brain volume and ventricles of each CT scan are automatically segmented by convolutional neural network (CNN) approaches. In the second step, we perform multi-atlas registration using the segmentation masks as guidance structures for the registration. We assume that a better alignment of the ventricles also leads to more precise propagation of all anatomical labels. Details are described in Section “Atlas Registration.” In our multi-atlas solution, CT scans are registered to an MRI brain atlas and with a mono-modal approach to three different pre-computed CT atlas images, such that we obtain four label images for each CT input. Finally, in the third step, we choose the label image with the highest ventricles Dice values. An overview of our approach is shown in Fig. 1. Registering a CT to all of the four atlas images and obtaining the final label image takes around 90 seconds on a system with NVIDIA GeForce RTX 2070 Super.Fig. 1Overview of our atlas registration approach: Step (1): automated segmentation of brain structures with CNNs, (2): multi-atlas registration, (3): create label imageFull size imageDataOur experiments are based on the publicly available data set provided by the Radiological Society of North America (RSNA) in collaboration with members of the American Society of Neuroradiology and MD.ai in the context of the RSNA challenge for intracranial hemorrhage detection [14]. The data set includes over 25,000 CT slices of the head, labeled with the type of hemorrhage, if present. We reconstructed 3D volumes from the 2D CT scans and selected a subset of 220 “normal” 3D volumes without hemorrhage. This corresponds to 220 subjects with one scan per subject. For

CT / Spiral CT of the Atlas misalignment and cervical spine

Approach combining pattern recognition and atlas registration. J Nucl Med 11(49):1875–1883. Google Scholar Muschelli J (2020) A publicly available, high resolution, unbiased CT brain template. Inf Process Manage Uncertain Knowl-Based Syst 1239:358. Google Scholar Vos PC, Išgum I, Biesbroek JM, Velthuis BK, Viergever MA (2013) Combined pixel classification and atlas-based segmentation of the ventricular system in brain CT Images. In: Medical Imaging 2013: Image Processing. vol. 8669. International Society for Optics and Photonics. SPIE. p. 460 – 465. Z, Qiu T, Huo L, Yu L, Shi H, Zhang Y, Wang H (2018) Deformable head atlas of chinese adults incorporating inter-subject anatomical variations. IEEE Access 6:51392–51400. Google Scholar Gao A, Chen M, Hu Q (2013) Non-rigid registration between brain CT images and MRI brain atlas by combining grayscale information, point correspondence on the midsaggital plane and brain surface matching. In: Proceedings of the 2nd International Conference on Computer Science and Electronics Engineering (ICCSEE 2013). Atlantis Press; 2013/03. p. 222–225. Z, Qiu T, Tian Y, Feng H, Zhang Y, Wang H (2021) Automated brain structures segmentation from PET/CT images based on landmark-constrained dual-modality atlas registration. Phys Med Biol 5:66. CAS Google Scholar Dubost F, de Bruijne M, Nardin M, Dalca AV, Donahue KL, Giese AK, Etherton MR, Wu O, Groot M, Niessen W, Vernooij M, Rost N, Schirmer MD (2020) Multi-atlas image registration of clinical data with automated quality assessment using ventricle segmentation. Med Image Anal 63:101698. PubMed PubMed Central Google Scholar Flanders AE, Prevedello LM, Shih G, Halabi SS, Kalpathy-Cramer J, Ball R, Mongan JT, Stein A, Kitamura FC, Lungren MP, Choudhary G, Cala L, Coelho L, Mogensen M, Moron F, Miller E, Ikuta I, Zohrabian V, McDonell O, Lincoln C, Shah L, Joyner D, Agarwal A, Lee RK, Nath J (2020) Construction of a machine learning dataset through collaboration: the RSNA 2019 brain CT hemorrhage challenge. Radiol: Artif Intell 5(2):e190211. Google Scholar Klein J, Wenzel M, Romberg D, Köhn A, Kohlmann P, Link F, Hänsch H, Dicken V, Stein R, Haase J, Schreiber A, Kasan R, Hahn H, Meine H (2020) QuantMed: Component-based deep learning platform for translational research. In: Medical imaging 2020: imaging informatics for healthcare, research, and applications. vol. 11318. International Society for Optics and Photonics. SPIE. p. 229 – 236. N, Landeau B, Papathanassiou D, Crivello F, Etard O, Delcroix N, Mazoyer B, Joliot M (2002) Automated anatomical labeling of activations in SPM using a macroscopic anatomical. CT-based Atlas of the Head and Neck The head and neck atlas was derived from a reduced resolution (256x256) CT MANIXdata set from the OSIRIX data sets. New Atlas Viewer. CT

Data from Head and Neck Cancer CT Atlas (Head-Neck-CT-Atlas)

\)94%), so that our multi-modal multi-atlas consisted of one MR and three CT atlases. The number of chosen CT images could easily be adapted to incorporate more variability. We are aware that typical atlas based approaches consist of a much larger number of atlases [23].However, we limit ourselves to three images because first, the total number of atlases should be balanced with the size of our data set. Using 10-20 images would mean that we are actually using 5–10% of the data set as atlases. While this might lead to a larger anatomical variety and thus better registration results, it would still bias the validity of the evaluation of our methodology. Second, we describe a bootstrap strategy to improve atlas registration using labels from a single MR atlas. The accuracy of bootstrapped CT labels is therefore highly dependent on the initial CT-MR registration quality, limiting the set of possible CT atlas candidates to those with very good CT registration quality. For this reason, we decided to use only the smallest possible number of three CT images for the evaluation of our approach, which is about 1.5% of the data.The multi-atlas registration for a new unseen CT image works as follows. We use the approach from Section “Registration approach” to independently register a new CT image to each of the four atlas images, such that we obtain four registration results. We use the Dice overlap of the ventricles as a quality criteria, as these labels are available for all atlas images as well as the new unseen CT image by our automatic CNN segmentation described in Section “Segmentation.” Thus, we are globally choosing the warped anatomical label image of the atlas with the highest ventricle Dice. We are aware that in multi-atlas scenarios it is common to use a local label fusion approach, namely majority voting [2,3,4]. We implemented this during development, but then focused on the previously described global Dice approach.Table 1 Segmentation results: Mean Dice coefficient and mean Hausdorff Distance (in mm), with related standard deviation. The test set consists of 22 volumesFull size table Fig. 2Automatic segmentation results for brain volume (left) and ventricles (right). The first and third images show the ground truth masks, whereas the second and fourth display the automatically generated masksFull size image Fig. 3Disease cases: Automatic segmentation results for brain volume and ventricles. The brain volume (in orange) and the left and right

Atlas and Anatomy of PET/MRI, PET/CT and SPECT/CT

For the thalamus label. So far, assuring the accurate and consistent segmentation of such structures exceeded our capacities.Furthermore, we found in our experiments that leveraging the single MR atlas with a bootstrapped multi-CT-MR atlas generally leads to much more robust and accurate results. However, instead of selecting CT images with the highest Dice values for our combined MR-CT multi-atlas, other criteria such as anatomical variability could be considered.ConclusionIn this paper, we presented a novel multi-atlas registration approach to obtain anatomical labels on CT scans using a standard MRI brain atlas. By using the detailed MRI information, we overcome the problem of creating an anatomical CT atlas. Furthermore, synthesizing MR images from CT, as found in the literature, is not needed as we directly use an MR atlas. Our method combines multi- and mono-modal registration and incorporates structure guidance with automatically segmented brain structures with CNNs. Thus, our registration guidance requires no manual interaction.As future work, further improving the CNN segmentation to simultaneously segment several brain structures will be investigated. Moreover, we plan to further validate our approach also on pathological brain scans. ReferencesLötjönen JM, Wolz R, Koikkalainen JR, Thurfjell L, Waldemar G, Soininen H, Rueckert D (2010) Fast and robust multi-atlas segmentation of brain magnetic resonance images. Neuroimage 2(49):2352–2365. Google Scholar Cabezas M, Oliver A, Lladó X, Freixenet J, Cuadra MB (2011) A review of atlas-based segmentation for magnetic resonance brain images. Comput Methods Programs Biomed 12:104. Google Scholar Asman AJ, Landman BA (2013) Non-local statistical label fusion for multi-atlas segmentation. Med Image Anal 2(17):194–208. Google Scholar Ding W, Li L, Zhuang X, Huang L (2020) Cross-modality multi-atlas segmentation using deep neural networks. In: Medical Image Computing and Computer Assisted Intervention–MICCAI 2020. Cham: Springer International Publishing. p. 233–242. N, Guerreiro F, McClelland J, Presles B, Modat M, Nill S, Dearnaley D, DeSouza N, Oelfke U, Knopf AC, Ourselin S, Cardoso MJ (2017) Iterative framework for the joint segmentation and CT synthesis of MR images: application to MRI-only radiotherapy treatment planning. Phys Med Biol 5(62):4237–4253. Google Scholar Roy S, Carass A, Jog A, Prince JL, Lee J (2014) MR to CT registration of brains using image synthesis. In: Medical Imaging 2014: image Processing. vol. 9034. International Society for Optics and Photonics. SPIE. p. 307 – 314. M, Steinke F, Scheel V, Charpiat G, Farquhar J, Aschoff P, Brady M, Schölkopf B, Pichler BJ (2008) MRI-based attenuation correction for PET/MRI: a novel

Atlas and Anatomy of PET/MRI, PET/CT and SPECT/CT:

Is segmented at a coarser level and does not contain such details. Therefore, the sulci cannot be accurately mapped by registration, resulting in larger Hausdorff values. An example is shown in Fig. 5a. Similarly, the distances for the ventricles are large when the subhorn of the lateral ventricles is well segmented in the ground truth, which is not the case in the atlas. Such a case is shown in Fig. 5b. It is well known that the Hausdorff distance is very sensitive to such outliers. For this purpose, we also provide more robust 95% Hausdorff distance (HD95) and average surface distance (AVD) [24, 25], confirming the good Dice values, see Table 2.RobustnessWe claim that adding CT images as auxiliary atlas images makes our overall approach more robust. To evaluate that, we compared the performance on the whole data set when using only the MR atlas versus using the multi-atlas with three CT images. The results are shown graphically in Fig. 6. We observed that the ventricle Dice is significantly improved for multiple cases when using the multi-atlas.In addition to using a multi-atlas, we also utilize structure guidance to improve the registration performance. Table 3 shows a comparison of the registration metrics when using no guidance, only mask alignment and the full proposed method with additional landmark alignment. We conduct this experiment on our test data set of 22 CT scans with manual segmentation masks. The ventricle Dice of the registration without guidance increased from 0.72 to 0.92 when using landmark and mask alignment.We also applied the Wilcoxon test for dependent samples. The difference between not using any guidance and guiding with masks is significant with p-values for Dice and Hausdorff distance of \(10^{-5}\) and \(7\times 10^{-4}\), respectively. The use of additional landmark guidance does not lead to a significant improvement over using only the masks (p-value of 0.061).Fig. 5Visualization of cases with high Hausdorff distance for a brain volume and b ventricles with ground truth in red/pink and atlas result in green. The arrows indicate the area with differencesFull size imageFig. 6Dice values for ventricles and brain volume when using only the MR atlas or the multi-atlas. The data set consists of all of the 217 CT scansFull size imageTable 3 Registration results without guidance vs guided by masks or masks and landmarks: Mean Dice coefficient and mean Hausdorff Distance (in mm), with related standard deviation. Data set includes 22

Development and Characterization of a Chest CT Atlas

Labeled cross-sectional anatomy of the mouse on micro-CT HOME vet-Anatomy Mouse - Whole body Labeled cross-sectional anatomy of the mouse on micro-CT Antoine MICHEAU, MD , Denis HOA, MD Antoine MICHEAU, MD : 2 Allée Charles Darwin, 34170 Castelnau-le-lez Denis HOA, MD : 2 Allée Charles Darwin, 34170 Castelnau-le-lez Publication date: May 30, 2018 | Last update: Dec 12, 2024 ISSN 2534-5087 These images of a normal female Swiss mouse have been acquired with a laboratory-based microCT system (nanoScan PET/CT Mediso (Budapest, Hungary) with an operation voltage of 50kVp and a 0,14mm pitch, with an intravenous injection of 2ml of Visipaque (320mg d’I/ml), at CERIMED (Centre Européen de Recherche en Imagerie Médicale (CERIMED), Université d’Aix-Marseille, Marseille, France).Images were post processed on Osirix Dicom viewer with transverse, sagittal and dorsal reconstructions.3D rendered images (bones and skin) were created on VGSTUDIOMAX by the Dr. Samuel Mérigeaud (Tridilogy, Montpellier, France). Labeled cross-sectional anatomy of the mouse on micro-CT Mouse - Thorax - CT: Heart, Trachea, Bronchi, Right lung, Left lung Mouse - CT: Liver, Left lateral lobe of liver, Caudate lobe, Right lobe of liver, Mouse - Anatomy atlas - CT: Digestive system, Urinary organs Mouse - Pelvis - Anatomy: Urogenital system, Urinary bladder, Uterus Cross-sectional anatomy of the mouse on high-resolution X-ray computed tomography (micro-CT): in vivo imaging on a murine model Anatomy of the laboratory mouse: in vivo imaging atlas on a high-resolution X-ray computed tomography (micro-CT) Mouse - 3D - Anatomy atlas: Skeleton, Bones, Osteology Mouse - Whole body (CT): 3D, Gross anatomy Quick Links Figures Figure 1 - Labeled cross-sectional anatomy of the mouse on micro-CT Figure 2 - Mouse - Thorax - CT: Heart, Trachea, Bronchi, Right lung, Left lung Figure 3 - Mouse - CT: Liver, Left lateral lobe of liver, Caudate lobe, Right lobe of liver, Figure 4 - Mouse - Anatomy atlas - CT: Digestive system, Urinary organs Figure 5 - Mouse - Pelvis - Anatomy: Urogenital system, Urinary bladder, Uterus Figure 6 - Cross-sectional anatomy of the mouse on high-resolution X-ray computed tomography (micro-CT): in vivo imaging on a murine model Figure 7 -. CT-based Atlas of the Head and Neck The head and neck atlas was derived from a reduced resolution (256x256) CT MANIXdata set from the OSIRIX data sets. New Atlas Viewer. CT CT Atlas; Diagnostic Strategies. Approach to H/N CT; Cases. Case Presentations; CT Atlas. Bony Skull Base. Axial CT; Coronal CT; Paranasal Sinuses. Axial CT; Petrous Temporal Bones.

CT atlas of the dog brain - PubMed

Chrisreddot3 Expert Cheater Posts: 463 Joined: Sun Mar 24, 2019 1:38 am Reputation: 83 Atlas Fallen Game Name: Atlas FallenGame Engine:FledgeGame Version:1.0Options Required: inf Health, inf momentum,resources multiplier,exp multiplier,fly or in jump....Steam Website: THX SunBeam Administration Posts: 4947 Joined: Sun Feb 04, 2018 7:16 pm Reputation: 4663 Re: Atlas Fallen Post by SunBeam » Thu Aug 10, 2023 1:07 am Had a quick look, interesting Engine (Fledge). Here's a super quick God Mode:Head to game folder (e.g.: G:\SteamLibrary\steamapps\common\Atlas Fallen).Go inside media-next folder.Open settings.json file with Notepad++.Add the following on top:Like so:Save file and reboot/start the game.The CVar is processed in Era::GameStateMain::OnUpdate() function, through the game's internal rapidjson reader/parser. The code will set appropriate CharacterStatsFlag to enabled state. As a consequence, you will get hit, see the blood splatter vignette, but health won't decrease; nor will you be able to heal.Kindly credit this post or link to FRF if you plan to leech this. It's not posted anywhere, you saw it here FIRST. Thank you very much!BR,Sun vinny2k Expert Cheater Posts: 233 Joined: Sun Jun 03, 2018 11:03 pm Reputation: 151 Re: Atlas Fallen Post by vinny2k » Thu Aug 10, 2023 2:57 am Quick Gold Script: Change Gold Amount and then Pick Up some to see the change Akira Table Makers Posts: 1302 Joined: Fri May 24, 2019 2:04 am Reputation: 1729 Re: Atlas Fallen Post by Akira » Thu Aug 10, 2023 4:44 am Code: Select allGame Name: Atlas FallenGame Process: AtlasFallen (DX12).exe | AtlasFallen (VK).exeGame Version: ?Game Engine: FledgeSavegame: C:\Program Files (x86)\Steam\userdata\\1230530\remoteSteam App ID: 1230530Cheat Engine: 7.5Game Website:If you like my work then please rate this post positive, it's easy, just hit the "Post reputation" button at the top right of the post.If you want to share the table then share the link to this post but do not upload this table anywhere else. AtlasFallen_1.1.CT Atlas Fallen (DX 12) | CE 7.5 | CT 1.1 (608.47 KiB) Downloaded 3403 times Scripts:-God Mode-Inf. Health-Inf. Momentum-No Momentum Consume-Set Field Of View-Set Run Speed-Multiple PointerThe Table can also be found in our Discord Server:[Link]If you have any questions or need help feel free to ask there, we have tables for a lot other games as well. AtlasFallen_1.0.CT Atlas Fallen (DX 12) | CE 7.5 | CT 1.0 (503.75 KiB) Downloaded 1887 times Last edited by Akira on Mon Aug 14, 2023 12:36 am, edited 4 times in total. SunBeam Administration Posts:

CT ATLAS Instagram photos and videos

\(M_\text {V}^T\) accordingly. Additionally let \(r_\ell ,t_\ell \in \mathbb {R}^3\), \(\ell \in {\mathcal {V}}\) be the centers of gravity (COGs) of the different ventricles, i.e., \(r_\text {LLV}\) is the COG of \(M^R_\text {LLV}\), \(t_\text {LLV}\) is the COG of \(M^T_\text {LLV}\), etc.For the registration, we then minimize the following objective function w.r.t. to deformation vector field y:$$\begin{aligned} \begin{aligned} J(R, T(y))&= \text {NGF}(R, T(y)) + \frac{\alpha }{2} \sum _{k=1}^3\Vert \Delta y_k\Vert _{L^2(\Omega )}^2 \\&\quad + \beta \int _{\Omega }^{}\psi (\det \nabla y(x))dx \\&\quad + \frac{\gamma }{2} \Big (\Vert M^T_\text {BP}(y)-M^R_\text {BP}\Vert ^2_{L^2(\Omega )} \\&\quad + \Vert M^T_\text {V}(y)-M^R_\text {V}\Vert ^2_{L^2(\Omega )}\Big ) \\&\quad + \frac{\delta }{2}\sum _{\ell \in {\mathcal {V}}} \Vert y(r_\ell )-t_\ell \Vert _2^2 \end{aligned} \end{aligned}$$ (1) with weights \(\alpha , \beta , \gamma , \delta > 0\), NGF distance measure$$\begin{aligned} \text {NGF}(R,T) = \frac{1}{2} \int _{\Omega }^{} 1 - \left( \frac{ \left\langle \nabla R(x),\nabla T(x)\right\rangle _{\varepsilon _R \varepsilon _T} }{ \Vert \nabla T(x)\Vert _{\varepsilon _T} \, \Vert \nabla R(x)\Vert _{\varepsilon _R} } \right) ^2 dx \end{aligned}$$ (2) where \( \left\langle x,y\right\rangle _{\varepsilon } := x^\top y+\varepsilon \), \(\Vert x\Vert _\varepsilon := \sqrt{\langle x,x\rangle _{\varepsilon ^2}}\) and \(\varepsilon _R,\varepsilon _T>0\) are the so-called edge-parameters controlling influence of noise in the images. The weights are fixed and were determined empirically. In addition to penalizing the second-order (Laplacian) derivatives by the so-called curvature regularization, we add an additional term penalizing the Jacobians of the deformation, respectively, volume changes with the function \(\psi (t)=(t-1)^2/t\) for \(t>0\) and \(\psi (t):=\infty \) for \(t\le 0\). Note that \(\psi (1)=0\) and \(\psi (t)=\psi (1/t)\) and thus volume growth or shrinkage are penalized symmetrically, and \(\psi (t)=\infty \) for \(\det \nabla y \le 0\) prevents local changes in the topology and thus unwanted mesh folds.The optimization is done by using a multi-level approach with L-BFGS.Multi-modal atlas registrationIn general, our approach builds on a single MR atlas that is transferred to CT as described before. However, to achieve better performance and coverage of anatomical variations, we bootstrap the MR atlas to a multi-modal MR-CT multi-atlas. To this end, all CT images in our data set (220 cases) were registered with the MR atlas intensity image and labels were propagated from MR to CT, so that we obtained a label image \(\text {CT}^\text {Label}\) for each CT scan. Afterward, we manually selected three CT images along with the propagated label images that had the highest ventricular Dice values (\(\ge. CT-based Atlas of the Head and Neck The head and neck atlas was derived from a reduced resolution (256x256) CT MANIXdata set from the OSIRIX data sets. New Atlas Viewer. CT CT Atlas; Diagnostic Strategies. Approach to H/N CT; Cases. Case Presentations; CT Atlas. Bony Skull Base. Axial CT; Coronal CT; Paranasal Sinuses. Axial CT; Petrous Temporal Bones.

Atlas and Anatomy of PET/CT - SpringerLink

Abstract Purpose Computed tomography (CT) is widely used to identify anomalies in brain tissues because their localization is important for diagnosis and therapy planning. Due to the insufficient soft tissue contrast of CT, the division of the brain into anatomical meaningful regions is challenging and is commonly done with magnetic resonance imaging (MRI). Methods We propose a multi-atlas registration approach to propagate anatomical information from a standard MRI brain atlas to CT scans. This translation will enable a detailed automated reporting of brain CT exams. We utilize masks of the lateral ventricles and the brain volume of CT images as adjuvant input to guide the registration process. Besides using manual annotations to test the registration in a first step, we then verify that convolutional neural networks (CNNs) are a reliable solution for automatically segmenting structures to enhance the registration process. Results The registration method obtains mean Dice values of 0.92 and 0.99 in brain ventricles and parenchyma on 22 healthy test cases when using manually segmented structures as guidance. When guiding with automatically segmented structures, the mean Dice values are 0.87 and 0.98, respectively. Conclusion Our registration approach is a fully automated solution to register MRI atlas images to CT scans and thus obtain detailed anatomical information. The proposed CNN segmentation method can be used to obtain masks of ventricles and brain volume which guide the registration. Similar content being viewed by others IntroductionComputed tomography (CT) imaging of the brain is widely used in radiology as it provides good image contrast to identify hemorrhages, cerebrovascular lesions and tumors. To determine the best treatment, each pathology has to be detected and localized as precisely as possible. CT is the modality of choice in acute patient care and is provided 24/7/365 in many hospitals. An emerging shortage of radiologists could be outweighed by precise automated CT exam reporting. The diagnosis and subsequent therapy depends on the anatomical localization, as symptoms and neurological disorders correspond with the affected brain area. Due to the poor soft tissue contrast of CT, precisely determining anatomical structures and differentiating brain areas is challenging. Magnetic resonance imaging (MRI) is used to highlight specific structures thanks to higher soft tissue contrast and the possibility of acquiring different MRI protocols. MRI mainly is the modality of neuroscience and elective clinical work-up. In daily practice, there is limited availability and there are controversies concerning the feasibility in acute symptomatic patients. Even