DirectDensity

White paper DirectDensity® Technical principles and implications for radiotherapy Dr. André Ritter, Dr. Nilesh Mistry, Siemens Healthineers siemens-healthineers.com/radiotherapy SIEMENS Healthineers White paper · DirectDensity® Introduction DirectDensity® CT simulators are used in radiotherapy with 2 primary While it may be be beneficial to design a more patient- goals: 1) creating an accurate geometric model of the specific scan protocol, it is often not practical, to patient, and 2) performing accurate dose calculations. optimize the imaging protocol within a busy radiotherapy department. It is challenging and time-consuming to For the task of dose calculations, there is a conversion optimize CT imaging protocols given the complex of HU values from the CT image into relative electron or relationships between kV, mAs, dose, contrast, and mass density. This task is typically performed in the clinic noise. To overcome these challenges, engineers at using a calibration curve that is subsequently used in the Siemens Healthcare have implemented functionalities treatment planning system for dose calculation. To such as CARE Dose 4D and CARE kV[3,4] which can potentially reduce the risk that a wrong calibration semi-automatically adapt tube current and voltage. curve is used during dose calculation, most clinics often However, managing several calibrations for different choose to image using a fixed tube voltage of 120 kV tube voltages within the treatment planning system, (for example). if supported at all, could hinder the workflow and be prone to errors. Thus, it is not always practical to However, these CT images are also used by physicians implement changes in kV settings. to define the target and organs at risk for treatment prescription. To get optimal contrast and image quality In response to these various challenges, Siemens that can help delineate structures, CT images could be Healthcare introduced the DirectDensity®1 algorithm, acquired using different kV settings based on patient which directly reconstructs images that can be size or on the presence of contrast media[1,2]. interpreted as showing relative electron density and mass density2 at any given kV setting. DirectDensity® For example, in the case of bariatric patients, who have eliminates the need for tube-voltage dependent system a higher X-ray attenuation, the output current of the calibration as it is a single linear relationship. This X-ray tube at lower or conventional kV settings may not creates scope for tube voltage adaption (e.g., CARE kV) be sufficient to produce the required contrast-to-noise to optimize imaging protocols based on patient anatomy. ratio. For these patients, higher X-ray tube voltages DirectDensity® is available on Siemens SOMATOM CT might be necessary. In the case of pediatric patients or scanners3 compatible with the software version syngo younger breast cancer patients, who might be CT VA30 (or higher). unnecessarily exposed to higher radiation doses with a conventional 120 kV scan protocol, the contrast-to-noise ratio in the images could be maintained by using a scan protocol with lower kV and hence potentially lowering the dose to these patients. [1] W. A. Kalender, P. Deak, M. Kellermeier, M. Straten and S. V. Vollmar. Application and patient size-dependent optimization of x-ray spectra for CT. Medical Physics, Vol. 36, pp. 993-1007, 2009. [2] C. Canstein and J. G. Korporaal. Reduction of contrast agent dose at low kV settings – White Paper. Siemens Healthcare, 2015. [3] K. Grant and B. Schmidt. Care kV – White Paper. Siemens Healthcare, 2011. [4] B. Schmidt, R. Raupach and T. Flohr. How to scan with CARE kV – User Guide. Siemens Healthcare, 2011. 1 DirectDensity® reconstruction is designed for use in Radiation Therapy Planning (RTP) only. DirectDensity® reconstruction is not intended to be used for diagnostic imaging. 2 As shown by measurements with a Gammex 467 Tissue Characterization Phantom comparing standard reconstruction and DirectDensity® reconstruction. Image value to relative electron/mass density conversion for the standard reconstruction was based on a two-linear-equations approach with individual calibration for each tube voltage. For DirectDensity® images, a single tube-voltage-independent linear conversion was used. 3 SOMATOM go.Up, SOMATOM go.All, SOMATOM go.Top, SOMATOM go.Sim, SOMATOM go.Open Pro and SOMATOM X.cite. 2 DirectDensity® · White paper Contents Introduction 2 The DirectDensity® algorithm 4 How to use DirectDensity® 6 Phantom validations 8 Clinical examples 12 References 14 3 White paper · DirectDensity® The DirectDensity® algorithm Image Space Input image Bone Bone image DirectDensity® detection image Forward projection Forward projection Back projection Effective water K thickness Electron and Material Effective bone mass density Attenuation decomposition thickness Synthesis projections Projection Space Figure 1: Flow chart of the DirectDensity® algorithm. The DirectDensity® algorithm reconstructs CT images Bone detection from single-energy CT acquisitions. The resulting CT values can be interpreted as relative electron and mass It is possible to approximate images of the basis material density. This is achieved by combining image-based bone bone by using a threshold on the input images. Voxels detection with a projection-based material decom- in the input images that have a CT value below this position. Synthetic projections of the relative electron threshold are assumed to contain water with variable and mass density are obtained using the two-material density depending on the CT value. For CT values above decomposition of water and bone. DirectDensity® this threshold, it is assumed that a voxel contains a images are finally reconstructed from the synthetic mixture of water and bone. projections of relative electron and mass density. The DirectDensity® algorithm can be combined with both The amount of bone in this voxel increases linearly with standard and iterative reconstruction. The processing CT value. The above is a reliable assumption for natural steps are detailed in the following para-graphs, and they body materials. Non-natural-body materials, for example are illustrated in the flow chart in Figure 1. metals or contrast agents, are a possible source of errors in this processing step. Input image and attenuation The attenuation in projection space, also called a sinogram, contains the attenuation information for each X-ray beam that travels through the patient from the single energy CT acquisition. The corresponding input images show the distribution of the attenuation coefficient as cross-sectional images. Input images can be obtained from attenuation with a back projection, and attenuation can be obtained from the input images with a forward projection. Both attenuation and input images serve as input for the algorithm. 4 DirectDensity® · White paper Material decomposition Synthesis In general, two-material decomposition from single- The sum of the products of the effective thicknesses energy CT is not possible. Nevertheless, projections of and the corresponding attenuation coefficients of both the bone image are used to obtain projections of the basis materials would yield the known attenuation line effective thickness of the basis material bone. A mono- integrals. Therefore, this is the inversion of the material tonic relation between effective thicknesses of both decomposition in projection space. However, the relative basis materials and the attenuation can be established electron and mass density of both basis materials is also using an exact physical attenuation model. This model known. Adding the products of either relative electron or incorporates the CT scanner and acquisition-specific mass density and effective thickness of water and bone parameters, for example tube voltage. This makes it for each projection yields line integrals of either relative possible to obtain projections of the effective thick- electron or mass density. This produces synthetic nesses of water from the known effective thicknesses projections, or in other words a sinogram of the of bone and the acquired attenuation. Thus, a complete distribution of relative electron and mass density. decomposition into both basis materials is possible. Effective water Density thickness of water Either electron or Synthesis + mass density line integral Effective bone Density thickness of bone Figure 2: Illustration of the synthesis step. Relative density mentioned in this figure can either be relative electron or relative mass density, depending on the selected DirectDensity® variant. In the synthesis steps, effective thickness of both materials is multiplied either by the corresponding relative electron or mass density. The sum of both products gives the value of a line integral either of relative electron or mass density, respectively. Calculated for each beam of the CT acquisition, this gives a sinogram of either the relative electron or mass density distribution. 5 White paper · DirectDensity® How to use DirectDensity® Enabling DirectDensity® syadmin LE Scan ! E Recon 2 . 11/25/1974 (45) 33465 best CT View&GO RTP Abdomen [facto ... im Recon Ranges Topogram Topogram Abdomen ------- cigm 512 SPR (32 cm) 0.01 0.5 may cm Sa 100 Exposure Time Abdomen Reconstruction Patient Marking Recon Favorites | General Recon | Recon&GO Imago Impression Recon Box Physio Recon | Auto Tasking Inline Options Close Scanv/Recon Wickwk2 120k FAST Window Kemel ADMIRE ADMIRE Strength IMAR Abdomen 1.50 Or 40 A3 Abdomen Off 40- On Off Body · Head Confirm patient position . Quantitative Special · Artificial 120 - pDOentity O :36 40 3.3s 6.55 › mDOenaty Figure 3: User interface on the CT acquisition workplace, where the Sd kernel can be selected in a recon job to enable the DirectDensity® algorithm. The DirectDensity® algorithm can be enabled during Typically, the treatment planning system will use a reconstruction by choosing dedicated reconstruction calibration curve to convert the CT value of images kernels. The standard reconstruction and iterative into relative electron and mass density. This is also true reconstruction can be found in the Sd36 kernel. when using DirectDensity® images. If the treatment Figure 3 shows the relevant user interface on the CT planning system expects certain calibration points to acquisition workplace. The resulting DICOM images of be given, then for DirectDensity® this can be achieved DirectDensity® reconstructions are provided in the same by calculating suitable calibration points, using the linear manner as for standard reconstructions. DirectDensity® relationship given above. These calibration points map images can be identified by the special kernel names in given CT values to corresponding relative electron and the relevant DICOM attributes, for example in the series mass densities. An example for the use with description. DirectDensity® is shown in Table 1. Image values and use in treatment planning systems Image value Relative electron and mass density DirectDensity® image values are provided in a HU-like -1024 0.0 scaling, but are proportional to the relative electron and mass density ρ. This means that image values can be -1000 0.0 converted to relative electron / mass density using +4000 5.0 the following equation: Table 1: Example of calibration points for use with DirectDensity®. For 12-bit DICOM images, there is usually a range of image values Image value between -1024 and -1000 that would result in negative electron ρ = +1 density. In this example, this scenario is prevented by adding an 1000 additional calibration point for -1024, which results in the range below -1000 being mapped to zero relative electron and mass density if necessary. 6 DirectDensity® · White paper Unlocking the potential of CT with DirectDensity® DirectDensity® Additional image reconstructions < ----- DirectDensity® Additional image reconstructions DirectDensity® V Sim&Go image + DICOM RT objects Treatment Treatment planning system planning system Without Sim&Go With Sim&Go • DirectDensity® reconstructions can be sent directly to • The scan protocol can be easily configured with: the treatment planning system for contouring and Primary image: DirectDensity® image for - dose calculation contouring and dose calculation • If any additional reconstructions are required for Up to 7 additional reconstructions for - contouring, manual contour propagation is needed advanced contouring • Preffered reconstructions can automatically be transfered to Sim&Go • Contours and primary image, i.e., the DirectDensity® image can be sent to the treatment planning system Figure 4: Illustration of workflow alternatives for DirectDensity®. Clinical images courtesy of: MAASTRO Clinic, Maastricht, Netherlands. The first and simplest workflow is the following: DirectDensity® has the potential to remove constraints DirectDensity® images can be sent directly to the of fixed tube voltage because DirectDensity® images treatment planning system and can be used for show reduced variations with regard to tube voltage1 contouring and dose calculation. and can be used for dose calculation1. In addition, other reconstructions with different settings, such as the However, there may be another workflow that unlocks kernel, can be performed. These additional the potential of DirectDensity® even further. Here, the reconstructions could be optimized to provide images goal is to have even better images that aid, for example, with potentially better contrast for better visualization contouring. In this second workflow, the full potential of that can aid the task of contouring. In summary, patient-specific CT acquisition is used, by adapting tube DirectDensity® helps to maintain consistency of dose current and voltage to the patient anatomy, and possibly calculation and could open new possibilities for improved using automatedexposure control mechanisms (e.g., visualization and greater confidence in contouring. CARE kV). 1 As shown by measurements with a Gammex 467 Tissue Characterization Phantom comparing standard reconstruction and DirectDensity reconstruction. HU value to relative electron density conversion for the standard reconstruction was based on a two-linear-equations approach with individual calibration for each tube voltage. For DirectDensity® images, a single tube-voltage-independent linear conversion was used. 7 White paper · DirectDensity® Validation of DirectDensity® on phantoms Standard reconstruction DirectDensity® reconstruction Figure 5: CT images of a Gammex 467 tissue characterization phantom obtained from the same 120 kV CT acquisition with a standard reconstruction and a DirectDensity® reconstruction. Window level: C = 40, W = 300. The Gammex 467 tissue characterization phantom For the standard reconstructions, a calibration was scanned at different tube voltages to validate the relationship using two linear equations was determined DirectDensity® algorithm. Figure 5 shows a standard for each tube voltage. With this relationship, the mean reconstruction and a DirectDensity® reconstruction CT value found for each Gammex 467 tissue substitute from the same acquisition carried out at a tube voltage was converted to relative electron and mass density of 120 kV. ρstandard. The single linear relationship given above was used to calculate relative electron and mass density CT value proportional to relative electron ρDirectDensity® from DirectDensity® reconstructions. Using and mass density all relative electron and mass densities calculated for Figure 6 shows the relationship between measured the Gammex 467 tissue substitutes, the root mean of mean image values for the Gammex 467 tissue the squared relative differences (RMS) of the relative substitutes and their true relative electron and mass electron and mass densities is calculated by the density. Comparing a standard reconstruction to following equation: DirectDensity®, the following observations can be made: With the standard reconstruction, the relationship √ ∑( ρDirectDensity® - ρstandard typically shows a different slope for CT values above and RMS = ) below a certain point between 0 and 100. Additionally, ρstandard the slope above this point depends on the tube voltage. This implies that a different calibration is needed for This RMS is a measure of relative deviations between each tube voltage. However, when using DirectDensity®, relative electron and mass densities obtained from the we observe a single linear relationship that is standard images and the DirectDensity® images. The independent of the tube voltage. This shows that one RMS values we found are shown in Figure 7 and are calibration curve in the TPS might be enough even if below 1.3% for all tube voltages. The deviations are the tube voltage used for acquisition is varied. comparable to the magnitude of statistical fluctuations or variations caused by changes in geometry (see Figure 8). 8 DirectDensity® · White paper Standard DirectDensity® 2000 2000 1500 1500 O 1000 1000 500 500 0 0 CT value CT value omega -- o -- o --- o- - - - - - -500 -500 -1000 -1000 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 relative electron density ρ relative electron density ρ O 70 kV O 80 kV O 100 kV O 120 kV O 140 kV (ρ–1)⋅1000 soft tissue bone Figure 6: Mean CT values of Gammex 467 tissue substitutes as a function of their true relative electron densities at several tube voltages. Additionally, the dashed line shows the expected relationship between DirectDensity® CT values and relative electron and mass density. 5.00 % 4.00 % 3.00 % RMS 2.00 % 1.27 % 1.19 % 0.95 % 1.00 % 0.78 % 0.64 % 0.00 % 70 kV 80 kV 100 kV 120 kV 140 kV Figure 7: Root mean of the squared relative differences (RMS) between relative electron densities obtained with a standard reconstruction and a DirectDensity® reconstruction. Relative electron densities of Gammex 467 tissue substitutes were measured. For standard reconstruction, a tube-voltage dependent two-linear-equation calibration was used. For DirectDensity®, the direct linear relation was used. 9 White paper · DirectDensity® Improved beam hardening correction 1.4 % 1.4 % 1.2 % 1.0 % 0.8 % 0.8 % 0.6 % 0.5 % Coefficient of variation 0.4 % 0.2 % 0.2 % Standard 0.0 % DirectDensity® Brain B-200 Figure 8: Coefficient of variation measured by varying the position of tissue substitute inserts in the Gammex 467 phantom at 120 kV. Thanks to the two-material attenuation model, A coefficient of variation for the CT value of each tissue DirectDensity® supports the reduction of beam substitute was calculated to get a measure of the hardening artifacts originating from dense, bone-like variability with regard to the arrangement. tissue. This can be observed in the reduction of streak artifacts between bone substitutes in the Gammex 467 For example, at 120 kV, the coefficient of variation of phantom, as can be seen in Figure 5. We observed this the brain substitute with a standard reconstruction was visible reduction of streak artifacts in the Gammex 467 0.8%. This declined to 0.5% with DirectDensity®. With phantom for all tested tube voltages. Moreover, bone-like tissue, an even greater reduction can be DirectDensity® supports increased CT value stability with observed. For example, for the B-200 low density bone regard to geometric variations1. This is also an intrinsic substitute, we observed a reduction in CT value effect of the improved two-material attenuation model variability from 1.4% to 0.2% with DirectDensity® at used in DirectDensity®. This was tested with several 120 kV. A reduction in the same order of magnitude repeated scans of the Gammex 467 phantom with varied was observed for all bone-like tissue substitutes. This arrangements of the tissue substitutes. is a result of the ability of DirectDensity® and its beam hardening correction to improve CT value stability with regard to geometric variations. 1 As shown by measurements with a Gammex 467 Tissue Characterization Phantom comparing standard reconstruction and DirectDensity® reconstruction. Image value to relative electron/mass density conversion for the standard reconstruction was based on a two-linear-equations approach with individual calibration for each tube voltage. For DirectDensity® images, a single tube-voltage-independent linear conversion was used. 10 DirectDensity® · White paper Dosimetric evaluation 70 kV 80 kV 100 kV 120 kV 140 kV 70 kV 100 % 80 kV 99,5 % 100 % 100 kV 99,4 % 99,6 % 100 % Comparison between: 120 kV 99,3 % 99,6 % 99,8 % 100 % DirectDensity® and standard image at the same tube voltage 140 kV 99,2 % 99,4 % 99,6 % 99,8 % 100 % DirectDensity® images at different tube voltages Figure 9: Passing rates resulting from a gamma analysis with criteria of 1 mm and 1%. Dose distributions for a 7-field MV photon treatment plan for an anthropomorphic thorax phantom were calculated. Standard images were converted to relative electron and mass density using a tube-voltage-dependent stoichiometric calibration. DirectDensity® images were converted using a tube-voltage-independent linear relation. Recently, Zhao et al.[6] conducted a study in which they Dosimetric differences were evaluated using gamma reconstructed standard CT images of the Gammex 467 analysis[5] with criteria of 1 mm and 1%. Pass rates equal tissue characterization phantom at 70 kV, 80 kV, 100 kV, to or above 99.9% were achieved by comparing standard 120 kV and 140 kV. They also reconstructed images and stoichiometric calibration with DirectDensity® images for these acquisitions. They used DirectDensity® at the same tube voltage. The same stoichiometric CT calibrations for the standard images passing rates were obtained in a similar study[7] which and used a single linear conversion as above for calculated a 7-field MV photon treatment plan on an DirectDensity® images. A 9-field MV photon treatment anthropomorphic thorax phantom. Passing rates for plan was created and the dose distributions were the anthropomorphic study are shown in Figure 9. calculated for all standard and DirectDensity® images. [5] D. A. Low, W. B. Harms, S. Mutic and J. A. Purdy. A technique for the quantitative evaluation of dose distributions. Medical Physics, Vol. 25, No. 5, pp. 656-661, 1998. [6] T. Zhao. Dosimetric Evaluation of Direct Electron Density Computed Tomography Images for Simplification of Treatment Planning Workflow. In: ASTRO’ 58th annual meeting, Boston, 2016. [7] T. Zhao, N. Mistry, R. Raupach, N. Huenemohr, A. Ritter, B. Sun, H. Li and S. Mutic. Evaluation of the Use of Direct Electron Density CT Images in Radiation Therapy. In: AAPM 58th annual meeting, Washington, DC, 2016. 11 White paper · DirectDensity® Clinical examples DirectDensity® DirectDensity® Standard Standard Figure 10: Figure 11: CT images of a head & neck case from the same 120 kV acquisition. CT images of a head & neck case from the same 100 kV acquisition Courtesy of MAASTRO Clinic, Maastricht, Netherlands. with contrast media. Courtesy of MAASTRO Clinic, Maastricht, Window level: C = 40, W=300. Netherlands. Window level: C = 40, W = 300. Figure 10 and Figure 11 show CT images of head and In DirectDensity® images, the influence of contrast neck scans at 120 kV and 100 kV of different patients, media on dose calculation might be reduced to an acquired with a SOMATOM Confidence® RT Pro. In both acceptable level, while in standard images the full figures, a DirectDensity® image is compared with the benefit of improved contrast media visualization at standard image from the same acquisition. In Figure lower tube voltages can be exploited. 10, a reduction of beam hardening artifacts with DirectDensity® can be observed, especially in the Figure 12 shows a visual comparison of dose area surrounding the teeth. distributions for the same beam configuration, calculated on DirectDensity® images and standard images. Visually, The scan in Figure 11 was performed with contrast the correspondence between both distributions is high. media at 100 kV, and shows, that DirectDensity® images This finding is supported by the dosimetric evaluation reduce the contrast of the contrast media. This indicates performed on phantoms. that when using contrast media, a workflow using DirectDensity® images for dose calculation and standard A detailed evaluation of the dosmietric impact of images for contouring might be beneficial. DirectDensity, when using different tube-voltage levels, can be found in [8]. [8] B. van der Heyden, M. Öllers, A. Ritter, F. Verhaegen, W. van Elmpt. Clinical evaluation of a novel CT image reconstruction algorithm for direct dose calculations. Physics and Imaging in Radiation Oncology, Vol. 2, pp. 11-16, 2017, doi: 10.1016/j.phro.2017.03.001. 12 DirectDensity® · White paper Dose Transport in HU Dose 25.39 Gy 25.38 Gy Transport in 25.4 Dose to medium 25.4 Dose to medium 22.50 Gy 22.50 Gy 20.00 Gy 20.00 Gy 17.50 Gy 17.50 Gy 15.00 Gy 15.00 Gy 12.50 Gy 12.50 Gy 10.00 Gy 10.00 Gy 7.50 Gy 7.50 Gy 5.00 Gy 5.00 Gy 50 Gy 3.60 Gy 0.00 Gy 0:00 Gy .................. - IEC 61217 - IEC 61217 - Head First-Supino Head First-Supine Y: -53.40 cm -231 HU Y: 53.40 cm F30s - Unapproved - Frontal -F30s F305 - Unapproved - Sagittal - F305 130s - Unapproved - Frontal - 130s 130s - Unapproved - Sagittal-130s Transport in medium Dose to med um Z: 15.10 GID X: 6.46 cm X: -6.46 cm Figure 12: Comparison of dose distributions (color wash) for the same treatment plan calculated on a DirectDensity® image series (Sd36, left) and a standard image series (Br36, right) within the treatment planning system (Eclipse, Varian Medical Systems). Courtesy of MAASTRO Clinic, Maastricht, Netherlands. Compatibility and limitations The current implementation of the DirectDensity® treatment planning systems require a kV-specific cali- algorithm can be used in combination with iterative bration curve to convert CT image values into relative reconstruction, iMAR1, HD FOV1,2 and respiratory-gated electron and mass density. A single linear relationship 4D CT1. The DirectDensity® algorithm is designed based that does not depend on the tube voltage of CT on the assumption that body materials are composed of acquisitions can unlock the unused potential of CT a mixture of water and bone with variable density. This is imaging in radiotherapy, as it removes the need for a reasonable assumption for natural body materials. scan protocols with fixed tube voltage. In addition, it Non-natural materials, for example metals and contrast would reduce the potential sources of errors that may be agents like iodine, will decrease accuracy and – as with introduced if the wrong TPS calibration curve is selected. conventional CT images – can potentially lead to image This new capability of varying tube voltage while using artifacts. It is technically possible to select DirectDensity® one calibration curve offers scope for designing scan kernels in reconstructions of Dual Energy scans, but the protocols that are more personalized to the patient. resulting DirectDensity® images are not suitable for At the same time, specific reconstructions provide Dual Energy post processing. images that may provide better visualization and better contrast for different tasks, especially contouring, in Conclusions radiotherapy planning. The DirectDensity® algorithm provides images in which Overall, we believe that the introduction of the CT values can be interpreted as showing relative DirectDensity® provides a new opportunity to easily electron and mass density. Currently, radiation therapy access images that are optimized for both dose calculation and contouring. 1 Some features may be optional on SOMATOM systems. 2 The image quality for the area outside the 50 cm standard scan field of view does not meet the image quality of the area inside the 50 cm standard scan field of view. Image artefacts may appear, depending on the patient setup and anatomy scanned. 13 White paper · DirectDensity® References [1] W. A. Kalender, P. Deak, M. Kellermeier, M. Straten and S. V. Vollmar. Application and patient size-dependent optimization of x-ray spectra for CT. Medical Physics, Vol. 36, pp. 993-1007, 2009. [2] C. Canstein and J. G. Korporaal. Reduction of contrast agent dose at low kV settings – White Paper. Siemens Healthcare, 2015. [3] K. Grant and B. Schmidt. Care kV – White Paper. Siemens Healthcare, 2011. [4] B. Schmidt, R. Raupach and T. Flohr. How to scan with CARE kV – User Guide. Siemens Healthcare, 2011. [5] D. A. Low, W. B. Harms, S. Mutic and J. A. Purdy. A technique for the quantitative evaluation of dose distributions. Medical Physics, Vol. 25, No. 5, pp. 656-661, 1998. [6] T. Zhao. Dosimetric Evaluation of Direct Electron Density Computed Tomography Images for Simplification of Treatment Planning Workflow. In: ASTRO’ 58th annual meeting, Boston, 2016. [7] T. Zhao, N. Mistry, R. Raupach, N. Huenemohr, A. Ritter, B. Sun, H. Li and S. Mutic. Evaluation of the Use of Direct Electron Density CT Images in Radiation Therapy. In: AAPM 58th annual meeting, Washington, DC, 2016. [8] B. van der Heyden, M. Öllers, A. Ritter, F. Verhaegen, W. van Elmpt. Clinical evaluation of a novel CT image reconstruction algorithm for direct dose calculations. Physics and Imaging in Radiation Oncology, Vol. 2, pp. 11-16, 2017, doi: 10.1016/j.phro.2017.03.001. 14 DirectDensity® · White paper 15 On account of certain regional limitations of sales rights and service availability, we cannot guarantee that all products included in this brochure are available through the Siemens sales organization worldwide. Availability and packaging may vary by country and is subject to change without prior notice. Some/All of the features and products described herein may not be available in the United States. The information in this document contains general technical descriptions of specifications and options as well as standard and optional features which do not always have to be present in individual cases. Siemens reserves the right to modify the design, packaging, specifications, and options described herein without prior notice. Please contact your local Siemens sales representative for the most current information. Note: Any technical data contained in this document may vary within defined tolerances. Original images always lose a certain amount of detail when reproduced. International version incl. U.S.. Siemens Healthineers Headquarters Siemens Healthcare GmbH Henkestr. 127 91052 Erlangen, Germany Phone: +49 9131 84-0 siemens-healthineers.com Published by Siemens Healthcare GmbH · online pdf · 8663 0420 · ©Siemens Healthcare GmbH, 2020