A feasibility study of using virtual noncontrast images synthesized from dual-energy computed tomography for radiotherapy treatment planning

Introduction

Computed tomography (CT) plays a fundamental role in modern radiotherapy treatment planning and involves the delineation of gross target volumes (GTVs) and organs at risk (OARs), as well as the conversion of CT values to relative electron density (ED) or stopping power ratio (SPR) for accurate dose calculation. Contrast-enhanced CT scans using iodine-containing contrast agents are commonly used to enhance target-to-tissue contrast (1,2), but the absence of contrast agents during treatment can lead to inaccurate CT-ED or SPR conversion and imprecise dose distribution (3,4). To address this issue, true noncontrast (TNC) images acquired before contrast injection are required for accurate radiotherapy treatment planning and dose calculation without interference from contrast agents (5,6).

However, the need for repeated scans poses several challenges, such as increased radiation exposure, longer scanning time, and reduced work efficiency. Moreover, patient motion between subsequent scans can cause the positions of GTVs and OARs to shift between TNC and true contrast-enhanced (TCE) images. The imperfect alignment by deformable image registration may impact the precision of delineation (7,8). Therefore, there is a pressing need to achieve accurate delineation and precise dose calculation within a single CT scan.

Dual-energy CT (DECT) is a promising imaging technique for oncologic imaging that simultaneously acquires two attenuation datasets from X-rays of different energy levels. According to the distinctive energy-dependence characteristic of different tissues, tissue can be distinguished and classified based on the photon absorption differences (9-14). Exploiting this capability, DECT can decompose an object into two different components with their content weights, producing basic material images, via a material decomposition algorithm (9). Based on the contents and attenuation coefficients from the National Institute of Standards and Technology (NIST) at various energy levels of the two basic materials, virtual monochromatic images (VMIs), within specific energy-level range, can be generated. Notably, these multiparameter images derived from DECT are in perfect geometric alignment. Popular research areas in the DECT in radiotherapy include SPR calculation via low- and high-energy VMIs or improvement of delineation precision using the low-energy VMIs (15-19). Additionally, DECT allows for the extraction of the iodine component via a material decomposition algorithm, resulting in tomographic images without iodine, known as virtual noncontrast (VNC) images (20-23). By using DECT to generate both TCE and VNC images in a single scan, clinicians can achieve precise delineation and treatment planning based on a single scan, thereby reducing the patient’s radiation dose and avoiding patient motion. In 2014, Yamada et al. conducted research on using VNC images for dose calculation, but their study was limited to a water phantom and an iodine phantom (24). In 2021, researchers attempted to incorporate clinical VNC images into treatment planning and dose calculation of abdominal and head tumors (25,26). Although previous studies have explored the use of VNC images for dose calculation, they have been limited to specific phantoms or investigations of small samples. A systematic investigation on the feasibility of photon dose calculation using clinical VNC images derived from DECT across various body regions and radiotherapy techniques is lacking.

In this study, we aimed to comprehensively investigate the equivalence of TNC and VNC images synthesized from DECT across various cancer species and different photon radiotherapy techniques to provide more evidence for the clinical applications of VC.

Methods

Study participants

This retrospective study included a total of 54 patients who underwent CT simulations in the department of radiotherapy of Peking Union Medical College Hospital between July 2023 and March 2024. The inclusion criteria were as follows: patients with both TNC images from single-energy CT scans and VNC images synthesized from the second-generation fast kilovoltage peak-switching spectral CT scans. Patients with incomplete image data or substantially different scan ranges between the two scans were excluded. The study cohort consisted of 20 patients with cervical tumors, 15 patients with esophageal cancer, 15 patients with laryngeal tumors and 4 patients with bone metastasis or hip replacement. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the institutional ethics board of the Peking Union Medical College Hospital (No. I-24PJ0295). Informed consent was obtained from all the patients.

Image acquisition

All participants were scanned using both the single-energy mode for TNC images and the gemstone spectral image (GSI) mode for VNC images with a second-generation fast kilovoltage peak-switching spectral CT device (Revolution, GE HealthCare, Chicago, IL, USA). For the single-energy mode, the following clinical precontrast scan protocols were performed: tube voltage, 140 kVp (pelvic tumor)/120 kVp [head-and-neck (H&N) or chest tumor]; tube current, 400 mA (H&N tumor)/325 mA (pelvic and chest tumor); and rotation time, 0.8 s. For the GSI mode, the postcontrast scan parameters were as follows: tube voltage, 80 and 140 kVp fast switching; tube current, 405 mA (H&N tumor)/400 mA (chest tumor)/520 mA (pelvic tumor); and rotation time, 0.8 s. The contrast agent was injected at 450 mgI/kg. VNC and 120 kVp-like images (TCE) were synthesized from data files of DECT. The slice thicknesses of both single-energy and dual-energy were 2.5 mm, and the matrix size was 512×512.

TNC-based treatment planning and dose recalculation

Figure 1 demonstrates the overall flowchart of this study. All selected patients underwent photon radiotherapy. Treatment plans based on TNC images (referred to as TNC plans) were optimized and calculated using the Eclipse version 15.6 treatment planning system (TPS; Varian, Palo Alto, CA, USA), according to the prescription dose and OAR constraints. Subsequently, the same beam parameters from the TNC plans were used to recalculate dose distributions based on VNC and TCE images without reoptimization (referred to as VNC plans and TCE plans, respectively). To conduct dose recalculation, TCE and VNC images were deformably registered to TNC images in advance (Plastimatch version 1.9.4, opensource software). The GTVs and clinical target volumes (CTVs) were delineated by radiation oncologists, and an isotropic margin of 3 to 5 mm was added to the CTV to create the planning target volumes (PTVs).

Figure 1 Summary of the study’s workflow. TNC, true noncontrast; TCE, true contrast-enhanced; VNC, virtual noncontrast; HU, Hounsfield unit; DVH, dose-volume histogram.

Different treatment techniques were used based on the tumor location: volume-modulated arc therapy (VMAT) for cervical cancer and laryngeal cancer and coplanar intensity-modulated radiotherapy (IMRT) for esophageal cancer. The prescription doses delivered to PTV were as follows: 36.0 Gy (1.8 Gy per fraction × 20 fractions) for cervical cancer and 50.0 Gy (2.0 Gy per fraction × 25 fractions) for esophageal cancer and laryngeal cancer. Typically, the prescription dose was delivered to cover 95% of the PTV.

Planning evaluation

To assess the feasibility of using VNC images for treatment planning and dose calculation, several metrics were employed to evaluate the dosimetric equivalence between TNC and VNC plans. These included the Hounsfield unit (HU) values of PTV and OARs, the gamma pass rate, and differences in dose-volume histogram (DVH) parameters.

For comparison of a dose distribution with a reference one in a quantitative manner via the calculation of the gamma value of each point—which is the minimum Euclidean distance in the dose-spatial domain—the agreement between evaluated and reference dose distributions can be calculated using two acceptance criteria: dose difference criterion (DDC) in percentage and distance-to-agreement (DTA) in millimeters. Gamma analysis produces gamma index values that are assigned to each point, with gamma index values ≤1 indicating passed and other values indicating failure. The percentage of passing points in the gamma distribution is referred to as the gamma pass rate (27-29). In this study, three pairs of DDC and DTA were considered: 3%/3 mm, 2%/2 mm, and 1%/1 mm (30-32). All gamma pass rates were calculated using SNC patient software (Sun Nuclear Corporation, Melbourne, FL, USA).

For the difference in DVH parameters, relative differences of DVH parameters were calculated for PTVs and OARs. The DVH parameters selected for PTVs were the dose covering 95% of the PTV (D95%), the dose covering 98% of the PTV (D98%), and the mean dose (Dmean). The specific DVH parameters of OARs for different types of cancer are presented in Table 1.

The Wilcoxon signed-rank test was used to compare dose differences, with statistical significance defined as P<0.05. Statistical analyses were performed using SPSS version 29.0 (IBM Corp., Armonk, NY, USA) (33).

Special circumstances

Due to the algorithm limitations of dual-energy CT, the generation of VNC images not only suppresses the iodine-based contrast agent components but also has a certain impact on substances with high density and high atomic number, such as bones. The cases analyzed in relevant feasibility studies are limited, and there is a lack of discussion on bone metastasis and tumors near bones. To encourage a more comprehensive discussion on the feasibility of VNC images in treatment planning, we specifically included four patients with bone metastasis or myeloma in this study. A dose of 30/40 Gy (3/4 Gy/fraction × 10 fractions) was delivered to the PTV using VMAT technology, and treatment plans were calculated with a sequentially integrated boost technique (34). For these four special cases, we mainly evaluated the dosimetric differences in PTV.

Results

Comparison of HU values of PTV and OARs

Table 2 demonstrates the HU value of various PTVs and OARs in TNC, TCE, and VNC images, as well as the HU differences between TCE/VNC images and TNC images (also in Figure 2). In TCE images, the iodine contrast agents had a significant impact on PTV, leading to the HU value of blood-rich PTV and OARs increasing to varying degrees (in esophageal cancers and laryngeal cancers). In VNC images, the HU values of PTV and most OARs were not significantly different from those in TNC images, indicating that the VNC algorithm can generate VNC images that are closer to the TNC situation, thereby avoiding differences in dose calculation. Meanwhile, in the cervical cancer group, the CT value differences between VNC and TNC images were larger than those differences between TCE and TNC images.

Figure 2 Bar graph of HU value differences of PTV and OARs in TCE and VNC images compared to the reference TNC image: (A) cervical cancers, (B) esophageal cancers, and (C) laryngeal cancers. HU, Hounsfield unit; TCE, true contrast-enhanced; TNC, true noncontrast; VNC, virtual noncontrast; PTV, planning target volume; SC, spinal cord; BS, brain stem; PL, parotid left; PR, parotid right; OAR, organ at risk.

Dosimetric comparison

For gamma pass rates, regardless of cancer type, all gamma pass rates between VNC and TNC plans were 1 under 2%/2 mm and 3%/3 mm metric (Table 3). The mean gamma pass rate of the esophageal cancer group in 1%/1 mm was the smallest, at 94.0%.

Table 4 presents the mean differences in DVH parameters between the TCE/VNC plans and TNC plans, along with the results of the statistical test. In the cervical cancer group, it was observed that although there were statistical differences in some DVH parameters in both the TCE/TNC plans and VNC/TNC plans, the differences were less than 0.2%, and the mean deviation between VNC and TNC plans was smaller than that between TCE and TNC plans. In particular, there were no significant differences in the 50% dose (D50%) of the small intestine and bladder between the VNC and TNC plans. Figure 3 shows the boxplots illustrating the relative deviations of dosimetric parameters for each approach. In the esophageal cancer group, there were no statistically significant differences in DVH parameters of the PTV and spinal cord in either the VNC/TNC plans or TCE/TNC plans. However, DVH parameters of heart and lung in TCE plans were statistically smaller than those of TNC, while there were no significant differences between the VNC and TNC plans. In the laryngeal cancer group, the results indicated that there were no significant differences in DVH parameters between the TCE/VNC plans and TNC plans for patients with H&N tumors.

Figure 3 Boxplots of the dose differences for the TCE and VNC plans compared to the reference TNC plan. The vertical axis shows the dose relative deviations as a percentage of the prescribed dose in the TNC plan. (A) Cervical cancers. (B) Esophageal cancers. (C) Laryngeal cancers. IQR, interquartile range; TCE, true contrast-enhanced; TNC, true noncontrast; VNC, virtual noncontrast; SI, small intestine; SC, spinal cord; BS, brain stem; PL, parotid left; PR, parotid right.

Figure 4 shows the dose distribution of VNC and TNC plans for patient 17 with cervical cancer, measured in centigray. It was found that the difference between the VNC plan and the TNC plan was relatively small.

Figure 4 Comparison between the VNC and TNC plans for patient 17. VNC, virtual noncontrast; TNC, true noncontrast.

To further illustrate the dosimetric differences, we calculated a difference plan between the VNC plan and TNC plan for patients 17, 25, and 30, respectively. Figure 5 demonstrates the dose volume histogram of these difference plans. The differences between the VNC plan and the TNC plan were smaller than those between the TCE plan and the TNC plan.

Figure 5 DVH of the difference plan between the TCE of the VNC plans and the corresponding TNC plans of (A,B) patient 17 (cervical), (C,D) patient 24 (esophageal), and (E,F) patient 39 (laryngeal). TCE, true contrast-enhanced; TNC, true noncontrast; PTV, planning target volume; VNC, virtual noncontrast; DVH, dose-volume histogram.

Special circumstances results

Table 5 compares the dosimetric indicators of PTV in four selected patients under different imaging conditions. From this, it can be observed that for patients with bone metastases and myeloma, compared to the TNC plan, the deviation of various dosimetric parameters of PTV in the VNC plan was within 0.3%, suggesting that the difference between the two was still relatively small, and the deviation was close to that of tumors in other areas. This indicates that although VNC images may cause certain bone suppression, they can still be used in the design of radiotherapy plans for bone metastases in terms of dosimetric indicators.

Discussion

In this study, with TNC serving as the reference, we systematically evaluated the feasibility of using VNC images synthesized from DECT for dose calculation and explored its dosimetric advantages compared to TCE images in photon radiotherapy. Our results showed excellent agreement in dosimetric parameters between VNC plans and conventionally created TNC plans for IMRT or VMAT techniques, with all gamma pass rates of 1 in 2 mm/2% and differences of DVH parameters within 1%. Furthermore, VNC plans demonstrated advantages in rich-blood OARs, such as the heart and small intestine.

For patients with cervical cancer, the dose distribution between TNC and TCE/VNC plans were equivalent according to gamma pass rates in general. However, from Table 3 and Figure 3, it can be observed that the doses of PTV and OARs in TCE plans were often statistically underestimated, while the doses of PTV and OARs in the VNC plans appear to be slightly overestimated. This difference is primarily attributable by the generation method of VNC images. In material decomposition process of DECT, voxels in the scanned object are treated as weighted by two basic substances, water and iodine (35). Due to the presence of substances such as calcium in bones and contrast agent in the small intestine, the material decomposition algorithm used in DECT assumes higher iodine content in these organs, resulting in lower densities in VNC images compared to actual values (36), which was also pointed out by Kaufmann et al. (37). Therefore, in the calculation of dose in VNC plans, the attenuation of photons when passing through the human body is smaller, leading to a higher dose in the PTV, while the dose in the TCE plan is lower due to the higher HU values. This finding is consistent with previous research by Noid et al. (25).

For patients with esophageal tumors, there was no significant difference in the DVH parameters between the VNC and TNC plans. This was mainly because, for patients with upper esophageal cancer, the PTVs are located on the anterior side of the body, and the beams in the IMRT plan primarily pass a short distance in the patient’s body to reach the PTV from the front. Moreover, there are fewer bones along the beam path, resulting in less impact on the dose of the PTV due to image features. Meanwhile, for OARs, the density changes of the heart between TCE and TNC images were greater due to the flow of iodine-based contrast agent; meanwhile, in VNC images, the absence of high-density substances resulted in a marginal HU difference between TNC and VNC images based on the VNC algorithm (Figure 2). Therefore, the dosimetric parameters of the heart in VNC plans were similar to those in TNC plans, while in TCE plans, the differences were statistically significant (Table 2). As for the spinal cord, significant dose differences were not observed regardless of the plan type, since the photons had already attenuated significantly by the time they reached the spinal cord.

Based on the analysis of dosimetric parameters for the three types of tumors mentioned above, the doses in VNC plans were close to those in TNC plans, the relative difference was less than 1%, and most indicators showed no significant differences. These findings provide preliminary evidence for the feasibility of using VNC images for treatment planning and dose calculation. In contrast, there were notable differences between the doses in TCE plans and TNC plans, particularly in pelvic tumors such as cervical cancer and rich-blood OARs.

Several important considerations should be noted. First, our study solely focused on photon dose calculation, and the use of proton or heavy ion radiotherapy in clinical practice requires further exploration in terms of dose calculation. Second, our research was limited to adult patients, and it is recognized that precision radiotherapy plays a pivotal role in treating pediatric patients, significantly impacting their posttreatment quality of life. Further research is encouraged to assess the applicability of our findings in pediatric populations. Furthermore, for patients with metal implants, the incorporating the metal artifact removal (algorithm and VNC derived from DECT can be further explored.

Additionally, only the second-generation fast kilovoltage peak-switching spectral CT was used in this study. As demonstrated by other articles, the HU values of VNC images can vary significantly due to the different technologies and DECT scanners, especially between dual-source DECT and dual-layer DECT or kilovoltage peak-switching DECT (35,38,39). In addition, the differences in VNC process between second-generation and third-generation spectral CT may also influence the generated VNC images. Therefore, a comprehensive understanding of these technical nuances is essential for optimizing the use of DECT in radiotherapy planning.

In this study, we mainly investigated the feasibility of VNC images for dose calculation. In the study, we used Plastimatch for deformation registration to eliminate the differences caused by organ motion. However, one major advantage of the VNC technique is that it can simultaneously obtain TCE and VNC images in a single scan, completely avoiding the influence of organ movement. The geometric benefits of VNC could be examined in future investigations.

Conclusions

In this study, we analyzed the images derived from DECT of 54 patients with various types of cancer. The dosimetric accuracy of TNC, TCE, and VNC images was evaluated for photon radiotherapy plans, with a focus on comparing the dosimetric parameters of PTVs and OARs across different plans. Our findings demonstrated that for the majority of PTVs and OARs, the deviation between dose indicators in the TNC and VNC plans was within 1%, with no statistically significant differences. This outcome provides preliminarily evidence supporting the feasibility of using VNC images for photon dose calculation in radiotherapy.

Our research suggests that the need for TNC CT scans during the simulation process can be eliminated, leading to improved clinical workflow and reduced radiation dose to patients. Furthermore, we specifically investigated the feasibility of using VNC images for patients with tumors located near bones, adding to the systematic analysis of VNC image utilization in treatment planning. These findings serve as a valuable reference for the future clinical implementation of related technologies.

Acknowledgments

Funding: This work was supported by National Key R&D Program of China, Ministry of Science and Technology of the People’s Republic of China (No. 2022YFC2404606).

Footnote

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-885/coif). S.W., Q.Z., L.Y., W.L., B.Z., B.Y., and J.Q. report that they received funding from the National Key R&D Program of China, Ministry of Science and Technology of the People’s Republic of China (No. 2022YFC2404606). M.G., J.D., and X.L. are employees of GE HealthCare. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by institutional ethics board of the Peking Union Medical College Hospital (No. I-24PJ0295). Informed consent was obtained from all the patients.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Lee AW, Ng WT, Pan JJ, Poh SS, Ahn YC, AlHussain H, et al. International guideline for the delineation of the clinical target volumes (CTV) for nasopharyngeal carcinoma. Radiother Oncol 2018;126:25-36. [Crossref] [PubMed]
2. Tummala P, Junaidi O, Agarwal B. Imaging of pancreatic cancer: An overview. J Gastrointest Oncol 2011;2:168-74. [PubMed]
3. Marin D, Boll DT, Mileto A, Nelson RC. State of the art: dual-energy CT of the abdomen. Radiology 2014;271:327-42. [Crossref] [PubMed]
4. Ramm U, Damrau M, Mose S, Manegold KH, Rahl CG, Böttcher HD. Influence of CT contrast agents on dose calculations in a 3D treatment planning system. Phys Med Biol 2001;46:2631-5. [Crossref] [PubMed]
5. Liu AJ, Vora N, Suh S, Liu A, Schultheiss TE, Wong JY. Effect of CT contrast on volumetric arc therapy planning (RapidArc and helical tomotherapy) for head and neck cancer. Med Dosim 2015;40:32-6. [Crossref] [PubMed]
6. Mei K, Ehn S, Oechsner M, Kopp FK, Pfeiffer D, Fingerle AA, Pfeiffer F, Combs SE, Wilkens JJ, Rummeny EJ, Noël PB. Dual-layer spectral computed tomography: measuring relative electron density. Eur Radiol Exp 2018;2:20. [Crossref] [PubMed]
7. Yang D, Chaudhari SR, Goddu SM, Pratt D, Khullar D, Deasy JO, El Naqa I. Deformable registration of abdominal kilovoltage treatment planning CT and tomotherapy daily megavoltage CT for treatment adaptation. Med Phys 2009;36:329-38. [Crossref] [PubMed]
8. Nenoff L, Matter M, Amaya EJ, Josipovic M, Knopf AC, Lomax AJ, Persson GF, Ribeiro CO, Visser S, Walser M, Weber DC, Zhang Y, Albertini F. Dosimetric influence of deformable image registration uncertainties on propagated structures for online daily adaptive proton therapy of lung cancer patients. Radiother Oncol 2021;159:136-43. [Crossref] [PubMed]
9. Alvarez RE, Macovski A. Energy-selective reconstructions in X-ray computerized tomography. Phys Med Biol 1976;21:733-44. [Crossref] [PubMed]
10. Su T, Kaftandjian V, Duvauchelle P, Zhu Y. Classification-based material decomposition method for photon counting-based spectral radiography: Application to plastic sorting. Nucl Instrum Methods Phys Res A 2020;960:163537. [Crossref]
11. Johnson TR, Krauss B, Sedlmair M, Grasruck M, Bruder H, Morhard D, Fink C, Weckbach S, Lenhard M, Schmidt B, Flohr T, Reiser MF, Becker CR. Material differentiation by dual energy CT: initial experience. Eur Radiol 2007;17:1510-7. [Crossref] [PubMed]
12. Petersilka M, Bruder H, Krauss B, Stierstorfer K, Flohr TG. Technical principles of dual source CT. Eur J Radiol 2008;68:362-8. [Crossref] [PubMed]
13. Fredenberg E. Spectral and dual-energy X-ray imaging for medical applications. Nucl Instrum Methods Phys Res A 2018;878:74-87. [Crossref]
14. Wang X, Shen H, Zhang J, Liu D, Tao J, Luo Y, Chen L, Long L, Huang J, Huang Y. Dual-Energy CT: A New Frontier in Oncology Imaging. Meta-Radiology 2023;1:100044. [Crossref]
15. Xu X, Xie Y, Zhao M, Yuan C, Wang L. Investigation on Accuracy of Stopping Power Ratio Prediction Based on Spectral CT. J Med Biol Eng 2022;42:845-52. [Crossref]
16. Zhu L, Du Y, Peng Y, Xiang X, Wang X. Investigation on the proton range uncertainty with spectral CT-based virtual monoenergetic images. J Appl Clin Med Phys 2023;24:e14062. [Crossref] [PubMed]
17. Taasti VT, Petersen JB, Muren LP, Thygesen J, Hansen DC. A robust empirical parametrization of proton stopping power using dual energy CT. Med Phys 2016;43:5547. [Crossref] [PubMed]
18. Lennartz S, Große Hokamp N, Zäske C, Zopfs D, Bratke G, Glauner A, Maintz D, Persigehl T, Chang DH, Hickethier T. Virtual monoenergetic images preserve diagnostic assessability in contrast media reduced abdominal spectral detector CT. Br J Radiol 2020;93:20200340. [Crossref] [PubMed]
19. Zopfs D, Große Hokamp N, Reimer R, Bratke G, Maintz D, Bruns C, Mallmann C, Persigehl T, Haneder S, Lennartz S. Value of spectral detector CT for pretherapeutic, locoregional assessment of esophageal cancer. Eur J Radiol 2021;134:109423. [Crossref] [PubMed]
20. Lundin M, Lidén M, Magnuson A, Mohammed AA, Geijer H, Andersson T, Persson A. Virtual non-contrast dual-energy CT compared to single-energy CT of the urinary tract: a prospective study. Acta Radiol 2012;53:689-94. [Crossref] [PubMed]
21. Shi L, Yan F, Pan Z, Liu B, Liu H, Wang B, Zhang H, Yu Y. Stomach virtual non-enhanced CT with second-generation, dual-energy CT: a preliminary study. PLoS One 2014;9:e112295. [Crossref] [PubMed]
22. Euler A, Parakh A, Falkowski AL, Manneck S, Dashti D, Krauss B, Szucs-Farkas Z, Schindera ST. Initial Results of a Single-Source Dual-Energy Computed Tomography Technique Using a Split-Filter: Assessment of Image Quality, Radiation Dose, and Accuracy of Dual-Energy Applications in an In Vitro and In Vivo Study. Invest Radiol 2016;51:491-8. [Crossref] [PubMed]
23. De Cecco CN, Darnell A, Macías N, Ayuso JR, Rodríguez S, Rimola J, Pagés M, García-Criado A, Rengo M, Laghi A, Ayuso C. Virtual unenhanced images of the abdomen with second-generation dual-source dual-energy computed tomography: image quality and liver lesion detection. Invest Radiol 2013;48:1-9. [Crossref] [PubMed]
24. Yamada S, Ueguchi T, Ogata T, Mizuno H, Ogihara R, Koizumi M, Shimazu T, Murase K, Ogawa K. Radiotherapy treatment planning with contrast-enhanced computed tomography: feasibility of dual-energy virtual unenhanced imaging for improved dose calculations. Radiat Oncol 2014;9:168. [Crossref] [PubMed]
25. Noid G, Schott D, Paulson E, Zhu J, Shah J, Li XA. Technical Note: Using virtual noncontrast images from dual-energy CT to eliminate the need of precontrast CT for x-ray radiation treatment planning of abdominal tumors†. Med Phys 2021;48:1365-71. [Crossref] [PubMed]
26. Ohira S, Komiyama R, Koike Y, Washio H, Kanayama N, Inui S, Ueda Y, Miyazaki M, Koizumi M, Teshima T. Dual-energy computed tomography image-based volumetric-modulated arc therapy planning for reducing the effect of contrast-enhanced agent on dose distributions. Med Dosim 2021;46:328-34. [Crossref] [PubMed]
27. Low DA, Harms WB, Mutic S, Purdy JA. A technique for the quantitative evaluation of dose distributions. Med Phys 1998;25:656-61. [Crossref] [PubMed]
28. Low DA, Dempsey JF. Evaluation of the gamma dose distribution comparison method. Med Phys 2003;30:2455-64. [Crossref] [PubMed]
29. Park JM, Kim JI, Park SY, Oh DH, Kim ST. Reliability of the gamma index analysis as a verification method of volumetric modulated arc therapy plans. Radiat Oncol 2018;13:175. [Crossref] [PubMed]
30. Van Dyk J, Barnett RB, Cygler JE, Shragge PC. Commissioning and quality assurance of treatment planning computers. Int J Radiat Oncol Biol Phys 1993;26:261-73. [Crossref] [PubMed]
31. Nelms BE, Simon JA. A survey on planar IMRT QA analysis. J Appl Clin Med Phys 2007;8:76-90. [Crossref] [PubMed]
32. Abolaban F, Zaman S, Cashmore J, Nisbet A, Clark CH. Changes in Patterns of Intensity-modulated Radiotherapy Verification and Quality Assurance in the UK. Clin Oncol (R Coll Radiol) 2016;28:e28-34. [Crossref] [PubMed]
33. Argyrous G. Statistics for research: With a guide to SPSS. 3rd edition, SAGE Publications, 2011:1-608.
34. Miyazaki M, Nishiyama K, Ueda Y, Ohira S, Tsujii K, Isono M, Masaoka A, Teshima T. Preliminary analysis of the sequential simultaneous integrated boost technique for intensity-modulated radiotherapy for head and neck cancers. J Radiat Res 2016;57:406-11. [Crossref] [PubMed]
35. De Cecco CN, Muscogiuri G, Schoepf UJ, Caruso D, Wichmann JL, Cannaò PM, Canstein C, Fuller SR, Snider L, Varga-Szemes A, Hardie AD. Virtual unenhanced imaging of the liver with third-generation dual-source dual-energy CT and advanced modeled iterative reconstruction. Eur J Radiol 2016;85:1257-64. [Crossref] [PubMed]
36. Mileto A, Mazziotti S, Gaeta M, Bottari A, Zimbaro F, Giardina C, Ascenti G. Pancreatic dual-source dual-energy CT: is it time to discard unenhanced imaging? Clin Radiol 2012;67:334-9. [Crossref] [PubMed]
37. Kaufmann S, Sauter A, Spira D, Gatidis S, Ketelsen D, Heuschmid M, Claussen CD, Thomas C. Tin-filter enhanced dual-energy-CT: image quality and accuracy of CT numbers in virtual noncontrast imaging. Acad Radiol 2013;20:596-603. [Crossref] [PubMed]
38. Lennartz S, Pisuchpen N, Parakh A, Cao J, Baliyan V, Sahani D, Hahn PF, Kambadakone A. Virtual Unenhanced Images: Qualitative and Quantitative Comparison Between Different Dual-Energy CT Scanners in a Patient and Phantom Study. Invest Radiol 2022;57:52-61. [Crossref] [PubMed]
39. Durieux P, Gevenois PA, Muylem AV, Howarth N, Keyzer C. Abdominal Attenuation Values on Virtual and True Unenhanced Images Obtained With Third-Generation Dual-Source Dual-Energy CT. AJR Am J Roentgenol 2018;210:1042-58. [Crossref] [PubMed]