Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

3D Printed Anatomical Model of a Rat for Medical Imaging

3D Printed Anatomical Model of a Rat for Medical Imaging Current Directions in Biomedical Engineering 2019;5(1):187-190 Miriam Exner*, Patryk Szwargulski, Peter Ludewig, Tobias Knopp and Matthias Graeser 3D Printed Anatomical Model of a Rat for Medical Imaging Abstract: For medical research, approximately 115 million 1 Introduction animals are needed every year. Rodents are used to test possible applications and procedures for the diagnosis of Medical research requires the use of laboratory animals for anatomical and physiological diseases. However, working toxological studies, pharmaceutical tests, tests of new with living animals increases the complexity of an medical procedures and showing the improvement of experiment. Accurate experimental planning is essential in imaging technologies within a real case scenario. The total order to fulfill the 3R rules (replace, reduce and refine). amount of used animals is unknown but estimated to be Especially in tracer-based imaging modalities, such as above 115 million per year worldwide [1]. 80 % of these are magnetic particle imaging (MPI), where only nanoparticles rodents (mice, rats, guinea-pigs) and rabbits [2]. Animal give a positive contrast, the anatomical structure of the rodent studies follow strict regulations to ensure the compliance is not visible without co-registration with another imaging with ethical rules. On the other hand, these rules increase the modality. This leads to problems in the experimental complexity as well as the necessary planning time of an planning, as parameters, such as field of view, rodent experiment. Following the 3R rules (replace, reduce and position and tracer concentration, have to be determined refine) [3] by using a phantom instead of the animal, the without visual feedback. In this work, a 3D CAD rat model is experiment is not only more ethical but also simpler presented, which can be used to improve the experiment regarding the necessary effort. This requires that the phantom planning and thus reduce the number of animals required. It shares the same features as a living animal. was determined using an anatomy atlas and 3D printed with In this work, a 3D CAD model is presented, which stereolithography. The resulting model contains the most mimics the main parts of the anatomy of a rat. With 14 %, important organs and vessels as hollow cavities. By filling rats are the second most frequently used experimental these with appropriate tracer materials, the phantom can be animals after mice [2]. The presented rat phantom can be used in different imaging modalities such as MPI, magnetic used to improve the planning of in vivo experiments and thus resonance imaging (MRI) or computed tomography (CT). In reduce the burden for animal studies. The model was created a first MPI measurement, the phantom was filled with in computer-aided design (CAD, SolidWorks, Dassault superparamagnetic nanoparticles. Finally, a successful Systèmes, Vélizy-Villacoublay, France) based on an anatomy visualization of all organs and vessels of the phantom was atlas of rats [4] and fabricated using additive manufacturing. possible. This enables the planning of the experiment and the It contains all essential organs and vessels, which are hollow optimization of experimental parameters for a region of and can be filled with different kinds of materials. This interest, where certain organs in a living animal are localized. allows an anatomically close distribution of the tracer to be Keywords: Rat Model, Magnetic Particle Imaging, 3D modeled. The model was evaluated using magnetic particle imaging (MPI). MPI images the distribution of Printing superparamagnetic iron oxide (SPIO) nanoparticles using https://doi.org/10.1515/cdbme-2019-0048 static and dynamic magnetic fields for excitation and spatial encoding. MPI combines a high spatial and temporal resolution as well as a high sensitivity. In addition, this ______ modality enables the performance of real-time in vivo *Corresponding author: Miriam Exner: Section for Biomedical imaging [5]. For the MPI measurements, the cavities of the Imaging, University Medical Center Eppendorf, Hamburg, phantom were filled with a nanoparticle contrast agent at Germany; Institute for Biomedical Imaging, Technical University different concentrations. The model can be used for the exact Hamburg, Hamburg, Germany, e-mail: miriam.exner@tuhh.de planning and design of the imaging sequences as well as the Patryk Szwargulski, Tobias Knopp, Matthias Graeser: Section for Biomedical Imaging, University Medical Center Eppendorf, position planning of the animal in the field of view (FOV). Hamburg, Germany; Institute for Biomedical Imaging, Technical Thus, the number of animals required for a medical study can University Hamburg, Hamburg, Germany be reduced. Beside the model itself, the first MPI images of Peter Ludewig: Department of Neurology, University Medical the model are presented. Center Eppendorf, Hamburg, Germany Open Access. © 2019 Miriam Exner et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 License. M. Exner et al., 3D Printed Anatomical Model of a Rat for Medical Imaging — 1 88 2 Material and Methods 2.1 Model Design The model creation was based on an anatomy atlas containing sagittal, coronal and transverse cut sections of rats [4]. Most of these rats are approximately 8 weeks old and weigh about 200 g. By measuring the dimensions and the shape of the main organs and vessels, an ordinary rat was modeled. Furthermore, the shape of some organs was modified based on MRI images of a mouse to better imitate the lying position of the rat during an experiment. The organs Figure 1: CAD model (a) and 3D printed and filled phantom (b). and vessels of the model were designed as hollow cavities The entire model measures 203.5 mm × 45 mm × 36 mm. It with a wall thickness of at least 1 mm and can be filled mimics an 8-week-old rat of about 200 g. individually. By selecting different tracer concentrations for The support skin (marked with ☆ in Figure 1a) has the the cavities, an organ-specific perfusion can be emulated. In shape of the lower half of the rat’s body and guarantees a order to facilitate the printing process and the subsequent reproducible positioning of the organs. It was fabricated with post-processing, the brain, the heart as well as the spleen and the less expensive Ultimaker 3 (Ultimaker B.V., the left kidney were created as individual parts separated Geldermalsen, Netherlands), a 3D printer based on fused from the other organs. In addition, the vessels were split into deposition modeling, as no waterproof printing is required. two parts, which can be joined after the printing process. The designed CAD model is shown in Figure 1a. The volumes of the individual organs and vessels are listed in Table 1. 2.3 MPI Measurement For the MPI measurement, the organs and vessels of the 2.2 3D Printing and Post-Processing model were filled with perimag (micromod, Rostock, Germany, prod. code: 102-00-132, plain surface) at different The organs and vessels of the model were fabricated with the concentrations (see Table 1). The printed model filled with stereolithography printer Form 2 (Formlabs GmbH, Berlin, tracer is shown in Figure 1b. In order to determine the tracer Germany), which features a UV laser spot size of 140 μm and concentrations of the individual organs and vessels, a steady a layer height between 25 and 100 μm. The non-magnetic state had to be assumed. When a tracer is injected into a clear resin V4 (Formlabs GmbH, Berlin, Germany) was used living rat, it is distributed differently throughout the body as printing material. As required for the phantom, this over time. After the tracer is injected via the vena cava, it technology enables waterproof results. Other printing passes the heart and after a tracer-dependent blood half-live, technologies, such as MultiJet Modeling, would also meet the it accumulates in certain organs such as the liver or requirements. The anatomical shapes of the phantom are spleen [6,7]. For the measurement, a state was assumed, in characterized by predominantly curved structures. This which the particles are distributed evenly in the blood results in a limitation of the printability of very small without accumulation. Since each organ is supplied with components such as the vessels. For this reason, the caudal blood to different degrees, the concentrations of the tracer parts of the aorta and the vena cava had to be extended to an were determined from the ratio of the blood volume inner diameter of 1 mm. Nevertheless, the model mostly contained in the respective organ [8] to the entire organ reflects the anatomical data in size and relative position. volume. The latter were derived from the CAD model. These After printing, it is necessary to flush out the resin values as well as the resulting tracer concentrations can be remaining in the model. This has to be done immediately and obtained from Table 1. The concentration in the heart was in a dark environment. Afterwards, the phantom has to be also used as the concentration for the vessels. post-cured. The model was further impregnated with Nano- The measurements were performed using a 3D field free Seal (JELN GmbH, Schwalmtal, Germany) to prevent the point preclinical MPI scanner (model: 1P MPI25/20 FF, absorption of water. Eventually, the vessels, which were Bruker, Ettlingen, Germany). For imaging, an excitation field printed in two parts, were stuck together via the connection amplitude of 12 mT in all three directions was applied. shown in Figure 1 (top right) using epoxy adhesive. M. Exner et al., 3D Printed Anatomical Model of a Rat for Medical Imagin g— 189 Further, a selection field with a gradient strength of -1 G = diag(-0.6, -0.6, 1.2) Tm was used. The resulting FOV measures 40 × 40 × 20 mm . The phantom was moved along the x- and z-directions inside the scanner bore to 21 positions (7 × 1 × 3). The resulting region of interest (ROI) of 220 × 40 × 40 mm covers the entire rat model. With this approach, issues connected to the usage of focus fields, such as field imperfections, could be neglected. Further, a single system matrix could be used for the reconstruction [9]. Due to the contrast limitations of the reconstruction, the large ratio of the highest to the lowest tracer concentration Figure 2: Reconstruction of all organs and vessels within a single cannot be reconstructed within a single image. Therefore, the image. The results are presented as maximum intensity kidneys, the brain and the vessels were reconstructed projection in coronal view. In (a) the unwindowed image is separately by dividing the reconstruction into three parts shown. In (b) the display range is set between 0 and 10 % of along the longitudinal axis of the rat. The 6 patches at the the signal maximum of the reconstructed image. first two x-positions were combined as well as the 9 patches at the middle three x-positions and 6 patches at the last two x-positions. The combined patches were reconstructed jointly solving one system of equations as proposed in [9]. An iterative regularized Kaczmarz algorithm [10] implemented in the open source programming language Julia was used as the solver. The reconstruction was performed with a relative regularization parameter of λ = 0.01, a signal-to-noise ratio (SNR) of 3 and 2 iterations. Figure 3: Division of the reconstruction into three parts. The results are presented as maximum intensity projection in Table 1: Overview of the organs and vessels with their volumes, coronal (top) and sagittal view (bottom). The head region (a), the contained blood volume and the tracer concentrations used for the thoracic and abdominal region (b) and the vessels (c) are the measurement. shown. compared to the kidneys and the small diameters of the Organ or Organ volume Mean blood Tracer vessels. In the image below (b), the display range was vessel in the phantom volume in a living concentration 3 3 -1 adjusted from 0 % to 10 % of the signal maximum of the [mm ] rat [mm ] [8] [µg(Fe) ml ] reconstructed image. All higher values were set to the 1) Kidney 495 450 2430 maximum value of the colorbar. While the brain and the 2) Spleen 318 140 1215 vessels are visible, the other organs can no longer be 3) Heart 1488 490 810 distinguished. Besides, the artifacts become more prominent. 4) Aorta 200 - 810 In Figure 3 the results of the divided reconstruction are 5) V. cava 590 - 810 presented. The images were normalized to the respective 6) Lungs 3610 660 486 maximum intensity. Additionally, the signal values below 7) Liver 14077 1660 270 10 %, 5 % and 25 % of the signal maximum of the respective 8) Brain 1576 41 74 part were set to zero. By separating the reconstruction, all organs and vessels could be successfully visualized. 3 Results 4 Discussion and Conclusion The reconstruction results of all organs and vessels within a The presented phantom is based on an anatomy atlas and single image are shown in Figure 2. In the unwindowed reflects the anatomy of the most important organs and vessels image (a), all organs of the thoracic and abdominal region of the rat. These were selected with a view to typical can be identified. However, the brain and the vessels are not experiments within the preclinical imaging field of MPI. In visible due to the low tracer concentration of the brain M. Exner et al., 3D Printed Anatomical Model of a Rat for Medical Imagin g— 190 the future, the model could be extended by further organs, most commonly used animals in medical research, its such as the gastrointestinal tract, or tumor models. anatomy was chosen for this first phantom. For large animals, The phantom can be used to evaluate medical the discussed problems in printing accuracy become less procedures, plan and train experiments and animal handling problematic, which in turn enables to design anatomical before in vivo experiments and to compare in vivo images to phantoms of larger animals in even higher accuracy. phantom images of the same structure. The burden for animal studies can be reduced by using the model to determine the Author Statement position of the entire rat as well as its organs and vessels in Research funding: The authors thankfully acknowledge the the FOV prior to an in vivo experiment and thus avoiding financial support by the DFG (grant number KN 1108/2-1) experiment failure due to a dispositioning. In addition, and the BMBF (grant number 05M16GKA). Conflict of experimental procedures, sequences and algorithms can be interest: Authors state no conflict of interest. Informed tested in a realistic setup without the regulatory overhead that consent: Informed consent has been obtained from all is necessary for animal studies. individuals included in this study. Ethical approval: The Comparing the organ volume in the phantom with the research related to human use complies with all the relevant mean blood volume in a living rat, the kidneys have a high national regulations, institutional policies and was performed blood content. Even if the kidneys of a living animal have a in accordance with the tenets of the Helsinki Declaration, and strong blood supply, the kidneys of the rats in the anatomy has been approved by the authors' institutional review board atlas used for the model creation are relatively small. or equivalent committee. However, the rat kidney can differ significantly in size [11]. The model was evaluated with MPI but could also be References adapted for other imaging modalities using other tracers such as iodine or gadolinium. Furthermore, the data acquired from [1] Taylor K, et al. Estimates for Worldwide Laboratory Animal MRI measurements could be overlaid on the MPI images to Use in 2005. Altern Lab Anim 2008;36:327-342. analyze the resolution behavior within the phantom. [2] European Commission. Commission Staff Working Document. Accompanying document to the Report from the Physiological questions cannot be represented with the Commission to the Council and the European Parliament. model. In addition, no dynamic experiments can be Seventh Report on the Statistics on the Number of Animals performed, since the model lacks a complete vascular system. used for Experimental and other Scientific Purposes in the Since MPI enables real-time imaging, a model that can plan Member States of the European Union. 2013. Part 1/5. and simulate dynamic in vivo experiments would enhance the [3] Russell WMS, Burch RL. The Principles of Humane Experimental Technique. London: Methuen; 1959. possibilities of the phantom. For this purpose, the model [4] Hayakawa T, Iwaki T. A Color Atlas of Sectional Anatomy of could be extended by a more complex vascular system that the Rat. Tokyo: Adosuri; 2008. connects the individual organs. Using a pump, a circulation [5] Gleich B, Weizenecker J. Tomographic imaging using the could be established in the model. nonlinear response of magnetic particles. Nature 2005;435:1214-1217. The presented model is scalable. Using a printer with a [6] Weizenecker J, et al. Three-dimensional real-time in vivo sufficiently high resolution, even mouse phantoms could be magnetic particle imaging. Phys Med Biol 2009;54:L1–L10. realized. 3D printing offers further possibilities to develop [7] Khandhar AP, et al. Evaluation of PEG-coated iron oxide even more realistic phantoms. Using recently presented nanoparticles as blood pool tracers for preclinical magnetic particle imaging. Nanoscale 2017;9:1299–1306. flexible and transparent stereolithography resin, the elastic [8] Oeff K, König A. Das Blutvolumen einiger Rattenorgane und properties of the vascular system could be reflected. Lately, it ihre Restblutmenge nach Entbluten bzw. Durchspülung. was shown that using human cells, even a 3D printed beating Bestimmung mit P -markierten Erythrocyten. Naunyn- heart can be realized [12]. Schmiedebergs Arch 1955;226:98-102. A 3D printed anatomical model of a rat was [9] Szwargulski P, et al. Efficient Joint Image Reconstruction of Multi-Patch Data Reusing a Single System Matrix in demonstrated. The phantom is a close fit to a real anatomical Magnetic Particle Imaging. IEEE Trans Med Imaging structure of a rat. Only small adaptions had to be made to 2019;38:932-944. ensure the printability of the phantom. All organs and vessels [10] Kaczmarz S. Angenäherte Auflösung von Systemen linearer of the phantom were successfully imaged using an MPI Gleichungen. Bull Internat Acad Polon Sci 1937;A35:355– tomograph. The model has proven to enhance the experiment [11] Treuting PM, et al. Comparative Anatomy and Histology. planning in advance to in vivo imaging. Due to the anatomy- London: Academic Press; 2018. like structure, many experiments in rodents can be evaluated [12] Noor N, et al. 3D Printing of Personalized Thick and in the phantom instead of a living animal. As rodents are the Perfusable Cardiac Patches and Hearts. Adv Sci 2019;1900344. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Current Directions in Biomedical Engineering de Gruyter

3D Printed Anatomical Model of a Rat for Medical Imaging

Loading next page...
 
/lp/de-gruyter/3d-printed-anatomical-model-of-a-rat-for-medical-imaging-7VPjVCTjZz

References (13)

Publisher
de Gruyter
Copyright
© 2019 by Walter de Gruyter Berlin/Boston
eISSN
2364-5504
DOI
10.1515/cdbme-2019-0048
Publisher site
See Article on Publisher Site

Abstract

Current Directions in Biomedical Engineering 2019;5(1):187-190 Miriam Exner*, Patryk Szwargulski, Peter Ludewig, Tobias Knopp and Matthias Graeser 3D Printed Anatomical Model of a Rat for Medical Imaging Abstract: For medical research, approximately 115 million 1 Introduction animals are needed every year. Rodents are used to test possible applications and procedures for the diagnosis of Medical research requires the use of laboratory animals for anatomical and physiological diseases. However, working toxological studies, pharmaceutical tests, tests of new with living animals increases the complexity of an medical procedures and showing the improvement of experiment. Accurate experimental planning is essential in imaging technologies within a real case scenario. The total order to fulfill the 3R rules (replace, reduce and refine). amount of used animals is unknown but estimated to be Especially in tracer-based imaging modalities, such as above 115 million per year worldwide [1]. 80 % of these are magnetic particle imaging (MPI), where only nanoparticles rodents (mice, rats, guinea-pigs) and rabbits [2]. Animal give a positive contrast, the anatomical structure of the rodent studies follow strict regulations to ensure the compliance is not visible without co-registration with another imaging with ethical rules. On the other hand, these rules increase the modality. This leads to problems in the experimental complexity as well as the necessary planning time of an planning, as parameters, such as field of view, rodent experiment. Following the 3R rules (replace, reduce and position and tracer concentration, have to be determined refine) [3] by using a phantom instead of the animal, the without visual feedback. In this work, a 3D CAD rat model is experiment is not only more ethical but also simpler presented, which can be used to improve the experiment regarding the necessary effort. This requires that the phantom planning and thus reduce the number of animals required. It shares the same features as a living animal. was determined using an anatomy atlas and 3D printed with In this work, a 3D CAD model is presented, which stereolithography. The resulting model contains the most mimics the main parts of the anatomy of a rat. With 14 %, important organs and vessels as hollow cavities. By filling rats are the second most frequently used experimental these with appropriate tracer materials, the phantom can be animals after mice [2]. The presented rat phantom can be used in different imaging modalities such as MPI, magnetic used to improve the planning of in vivo experiments and thus resonance imaging (MRI) or computed tomography (CT). In reduce the burden for animal studies. The model was created a first MPI measurement, the phantom was filled with in computer-aided design (CAD, SolidWorks, Dassault superparamagnetic nanoparticles. Finally, a successful Systèmes, Vélizy-Villacoublay, France) based on an anatomy visualization of all organs and vessels of the phantom was atlas of rats [4] and fabricated using additive manufacturing. possible. This enables the planning of the experiment and the It contains all essential organs and vessels, which are hollow optimization of experimental parameters for a region of and can be filled with different kinds of materials. This interest, where certain organs in a living animal are localized. allows an anatomically close distribution of the tracer to be Keywords: Rat Model, Magnetic Particle Imaging, 3D modeled. The model was evaluated using magnetic particle imaging (MPI). MPI images the distribution of Printing superparamagnetic iron oxide (SPIO) nanoparticles using https://doi.org/10.1515/cdbme-2019-0048 static and dynamic magnetic fields for excitation and spatial encoding. MPI combines a high spatial and temporal resolution as well as a high sensitivity. In addition, this ______ modality enables the performance of real-time in vivo *Corresponding author: Miriam Exner: Section for Biomedical imaging [5]. For the MPI measurements, the cavities of the Imaging, University Medical Center Eppendorf, Hamburg, phantom were filled with a nanoparticle contrast agent at Germany; Institute for Biomedical Imaging, Technical University different concentrations. The model can be used for the exact Hamburg, Hamburg, Germany, e-mail: miriam.exner@tuhh.de planning and design of the imaging sequences as well as the Patryk Szwargulski, Tobias Knopp, Matthias Graeser: Section for Biomedical Imaging, University Medical Center Eppendorf, position planning of the animal in the field of view (FOV). Hamburg, Germany; Institute for Biomedical Imaging, Technical Thus, the number of animals required for a medical study can University Hamburg, Hamburg, Germany be reduced. Beside the model itself, the first MPI images of Peter Ludewig: Department of Neurology, University Medical the model are presented. Center Eppendorf, Hamburg, Germany Open Access. © 2019 Miriam Exner et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 License. M. Exner et al., 3D Printed Anatomical Model of a Rat for Medical Imaging — 1 88 2 Material and Methods 2.1 Model Design The model creation was based on an anatomy atlas containing sagittal, coronal and transverse cut sections of rats [4]. Most of these rats are approximately 8 weeks old and weigh about 200 g. By measuring the dimensions and the shape of the main organs and vessels, an ordinary rat was modeled. Furthermore, the shape of some organs was modified based on MRI images of a mouse to better imitate the lying position of the rat during an experiment. The organs Figure 1: CAD model (a) and 3D printed and filled phantom (b). and vessels of the model were designed as hollow cavities The entire model measures 203.5 mm × 45 mm × 36 mm. It with a wall thickness of at least 1 mm and can be filled mimics an 8-week-old rat of about 200 g. individually. By selecting different tracer concentrations for The support skin (marked with ☆ in Figure 1a) has the the cavities, an organ-specific perfusion can be emulated. In shape of the lower half of the rat’s body and guarantees a order to facilitate the printing process and the subsequent reproducible positioning of the organs. It was fabricated with post-processing, the brain, the heart as well as the spleen and the less expensive Ultimaker 3 (Ultimaker B.V., the left kidney were created as individual parts separated Geldermalsen, Netherlands), a 3D printer based on fused from the other organs. In addition, the vessels were split into deposition modeling, as no waterproof printing is required. two parts, which can be joined after the printing process. The designed CAD model is shown in Figure 1a. The volumes of the individual organs and vessels are listed in Table 1. 2.3 MPI Measurement For the MPI measurement, the organs and vessels of the 2.2 3D Printing and Post-Processing model were filled with perimag (micromod, Rostock, Germany, prod. code: 102-00-132, plain surface) at different The organs and vessels of the model were fabricated with the concentrations (see Table 1). The printed model filled with stereolithography printer Form 2 (Formlabs GmbH, Berlin, tracer is shown in Figure 1b. In order to determine the tracer Germany), which features a UV laser spot size of 140 μm and concentrations of the individual organs and vessels, a steady a layer height between 25 and 100 μm. The non-magnetic state had to be assumed. When a tracer is injected into a clear resin V4 (Formlabs GmbH, Berlin, Germany) was used living rat, it is distributed differently throughout the body as printing material. As required for the phantom, this over time. After the tracer is injected via the vena cava, it technology enables waterproof results. Other printing passes the heart and after a tracer-dependent blood half-live, technologies, such as MultiJet Modeling, would also meet the it accumulates in certain organs such as the liver or requirements. The anatomical shapes of the phantom are spleen [6,7]. For the measurement, a state was assumed, in characterized by predominantly curved structures. This which the particles are distributed evenly in the blood results in a limitation of the printability of very small without accumulation. Since each organ is supplied with components such as the vessels. For this reason, the caudal blood to different degrees, the concentrations of the tracer parts of the aorta and the vena cava had to be extended to an were determined from the ratio of the blood volume inner diameter of 1 mm. Nevertheless, the model mostly contained in the respective organ [8] to the entire organ reflects the anatomical data in size and relative position. volume. The latter were derived from the CAD model. These After printing, it is necessary to flush out the resin values as well as the resulting tracer concentrations can be remaining in the model. This has to be done immediately and obtained from Table 1. The concentration in the heart was in a dark environment. Afterwards, the phantom has to be also used as the concentration for the vessels. post-cured. The model was further impregnated with Nano- The measurements were performed using a 3D field free Seal (JELN GmbH, Schwalmtal, Germany) to prevent the point preclinical MPI scanner (model: 1P MPI25/20 FF, absorption of water. Eventually, the vessels, which were Bruker, Ettlingen, Germany). For imaging, an excitation field printed in two parts, were stuck together via the connection amplitude of 12 mT in all three directions was applied. shown in Figure 1 (top right) using epoxy adhesive. M. Exner et al., 3D Printed Anatomical Model of a Rat for Medical Imagin g— 189 Further, a selection field with a gradient strength of -1 G = diag(-0.6, -0.6, 1.2) Tm was used. The resulting FOV measures 40 × 40 × 20 mm . The phantom was moved along the x- and z-directions inside the scanner bore to 21 positions (7 × 1 × 3). The resulting region of interest (ROI) of 220 × 40 × 40 mm covers the entire rat model. With this approach, issues connected to the usage of focus fields, such as field imperfections, could be neglected. Further, a single system matrix could be used for the reconstruction [9]. Due to the contrast limitations of the reconstruction, the large ratio of the highest to the lowest tracer concentration Figure 2: Reconstruction of all organs and vessels within a single cannot be reconstructed within a single image. Therefore, the image. The results are presented as maximum intensity kidneys, the brain and the vessels were reconstructed projection in coronal view. In (a) the unwindowed image is separately by dividing the reconstruction into three parts shown. In (b) the display range is set between 0 and 10 % of along the longitudinal axis of the rat. The 6 patches at the the signal maximum of the reconstructed image. first two x-positions were combined as well as the 9 patches at the middle three x-positions and 6 patches at the last two x-positions. The combined patches were reconstructed jointly solving one system of equations as proposed in [9]. An iterative regularized Kaczmarz algorithm [10] implemented in the open source programming language Julia was used as the solver. The reconstruction was performed with a relative regularization parameter of λ = 0.01, a signal-to-noise ratio (SNR) of 3 and 2 iterations. Figure 3: Division of the reconstruction into three parts. The results are presented as maximum intensity projection in Table 1: Overview of the organs and vessels with their volumes, coronal (top) and sagittal view (bottom). The head region (a), the contained blood volume and the tracer concentrations used for the thoracic and abdominal region (b) and the vessels (c) are the measurement. shown. compared to the kidneys and the small diameters of the Organ or Organ volume Mean blood Tracer vessels. In the image below (b), the display range was vessel in the phantom volume in a living concentration 3 3 -1 adjusted from 0 % to 10 % of the signal maximum of the [mm ] rat [mm ] [8] [µg(Fe) ml ] reconstructed image. All higher values were set to the 1) Kidney 495 450 2430 maximum value of the colorbar. While the brain and the 2) Spleen 318 140 1215 vessels are visible, the other organs can no longer be 3) Heart 1488 490 810 distinguished. Besides, the artifacts become more prominent. 4) Aorta 200 - 810 In Figure 3 the results of the divided reconstruction are 5) V. cava 590 - 810 presented. The images were normalized to the respective 6) Lungs 3610 660 486 maximum intensity. Additionally, the signal values below 7) Liver 14077 1660 270 10 %, 5 % and 25 % of the signal maximum of the respective 8) Brain 1576 41 74 part were set to zero. By separating the reconstruction, all organs and vessels could be successfully visualized. 3 Results 4 Discussion and Conclusion The reconstruction results of all organs and vessels within a The presented phantom is based on an anatomy atlas and single image are shown in Figure 2. In the unwindowed reflects the anatomy of the most important organs and vessels image (a), all organs of the thoracic and abdominal region of the rat. These were selected with a view to typical can be identified. However, the brain and the vessels are not experiments within the preclinical imaging field of MPI. In visible due to the low tracer concentration of the brain M. Exner et al., 3D Printed Anatomical Model of a Rat for Medical Imagin g— 190 the future, the model could be extended by further organs, most commonly used animals in medical research, its such as the gastrointestinal tract, or tumor models. anatomy was chosen for this first phantom. For large animals, The phantom can be used to evaluate medical the discussed problems in printing accuracy become less procedures, plan and train experiments and animal handling problematic, which in turn enables to design anatomical before in vivo experiments and to compare in vivo images to phantoms of larger animals in even higher accuracy. phantom images of the same structure. The burden for animal studies can be reduced by using the model to determine the Author Statement position of the entire rat as well as its organs and vessels in Research funding: The authors thankfully acknowledge the the FOV prior to an in vivo experiment and thus avoiding financial support by the DFG (grant number KN 1108/2-1) experiment failure due to a dispositioning. In addition, and the BMBF (grant number 05M16GKA). Conflict of experimental procedures, sequences and algorithms can be interest: Authors state no conflict of interest. Informed tested in a realistic setup without the regulatory overhead that consent: Informed consent has been obtained from all is necessary for animal studies. individuals included in this study. Ethical approval: The Comparing the organ volume in the phantom with the research related to human use complies with all the relevant mean blood volume in a living rat, the kidneys have a high national regulations, institutional policies and was performed blood content. Even if the kidneys of a living animal have a in accordance with the tenets of the Helsinki Declaration, and strong blood supply, the kidneys of the rats in the anatomy has been approved by the authors' institutional review board atlas used for the model creation are relatively small. or equivalent committee. However, the rat kidney can differ significantly in size [11]. The model was evaluated with MPI but could also be References adapted for other imaging modalities using other tracers such as iodine or gadolinium. Furthermore, the data acquired from [1] Taylor K, et al. Estimates for Worldwide Laboratory Animal MRI measurements could be overlaid on the MPI images to Use in 2005. Altern Lab Anim 2008;36:327-342. analyze the resolution behavior within the phantom. [2] European Commission. Commission Staff Working Document. Accompanying document to the Report from the Physiological questions cannot be represented with the Commission to the Council and the European Parliament. model. In addition, no dynamic experiments can be Seventh Report on the Statistics on the Number of Animals performed, since the model lacks a complete vascular system. used for Experimental and other Scientific Purposes in the Since MPI enables real-time imaging, a model that can plan Member States of the European Union. 2013. Part 1/5. and simulate dynamic in vivo experiments would enhance the [3] Russell WMS, Burch RL. The Principles of Humane Experimental Technique. London: Methuen; 1959. possibilities of the phantom. For this purpose, the model [4] Hayakawa T, Iwaki T. A Color Atlas of Sectional Anatomy of could be extended by a more complex vascular system that the Rat. Tokyo: Adosuri; 2008. connects the individual organs. Using a pump, a circulation [5] Gleich B, Weizenecker J. Tomographic imaging using the could be established in the model. nonlinear response of magnetic particles. Nature 2005;435:1214-1217. The presented model is scalable. Using a printer with a [6] Weizenecker J, et al. Three-dimensional real-time in vivo sufficiently high resolution, even mouse phantoms could be magnetic particle imaging. Phys Med Biol 2009;54:L1–L10. realized. 3D printing offers further possibilities to develop [7] Khandhar AP, et al. Evaluation of PEG-coated iron oxide even more realistic phantoms. Using recently presented nanoparticles as blood pool tracers for preclinical magnetic particle imaging. Nanoscale 2017;9:1299–1306. flexible and transparent stereolithography resin, the elastic [8] Oeff K, König A. Das Blutvolumen einiger Rattenorgane und properties of the vascular system could be reflected. Lately, it ihre Restblutmenge nach Entbluten bzw. Durchspülung. was shown that using human cells, even a 3D printed beating Bestimmung mit P -markierten Erythrocyten. Naunyn- heart can be realized [12]. Schmiedebergs Arch 1955;226:98-102. A 3D printed anatomical model of a rat was [9] Szwargulski P, et al. Efficient Joint Image Reconstruction of Multi-Patch Data Reusing a Single System Matrix in demonstrated. The phantom is a close fit to a real anatomical Magnetic Particle Imaging. IEEE Trans Med Imaging structure of a rat. Only small adaptions had to be made to 2019;38:932-944. ensure the printability of the phantom. All organs and vessels [10] Kaczmarz S. Angenäherte Auflösung von Systemen linearer of the phantom were successfully imaged using an MPI Gleichungen. Bull Internat Acad Polon Sci 1937;A35:355– tomograph. The model has proven to enhance the experiment [11] Treuting PM, et al. Comparative Anatomy and Histology. planning in advance to in vivo imaging. Due to the anatomy- London: Academic Press; 2018. like structure, many experiments in rodents can be evaluated [12] Noor N, et al. 3D Printing of Personalized Thick and in the phantom instead of a living animal. As rodents are the Perfusable Cardiac Patches and Hearts. Adv Sci 2019;1900344.

Journal

Current Directions in Biomedical Engineeringde Gruyter

Published: Sep 1, 2019

Keywords: Rat Model; Magnetic Particle Imaging; 3D Printing

There are no references for this article.