Preliminary study of clinical application on IMRT three‐dimensional dose verification‐based EPID system

Abstract The three‐dimensional dose (3D) distribution of intensity‐modulated radiation therapy (IMRT) was verified based on electronic portal imaging devices (EPIDs), and the results were analyzed. Thirty IMRT plans of different lesions were selected for 3D EPID‐based dose verification. The gamma passing rates of the 3D dose verification‐based EPID system (Edose, Version 3.01, Raydose, Guangdong, China) and Delta4 measurements were then compared with treatment planning system (TPS) calculations using global gamma criteria of 5%/3 mm, 3%/3 mm, and 2%/2 mm. Furthermore, the dose–volume histograms (DVHs) for planning target volumes (PTVs) as well as organs at risk (OARs) were analyzed using Edose. For dose verification of the 30 treatment plans, the average gamma passing rates of Edose reconstructions under the gamma criteria of 5%/3 mm, 3%/3 mm, and 2%/2 mm were (98.58 ± 0.93)%, (95.67 ± 1.97)%, and (83.13 ± 4.53)%, respectively, whereas the Delta4 measurement results were (99.14% ± 1.16)%, (95.81% ± 2.88)%, and (84.74% ± 7.00)%, respectively. The dose differences between Edose reconstructions and TPS calculations were within 3% for D95%, D98%, and Dmean in each PTV, with the exception that the D98% of the PTV‐clinical target volume (CTV) in esophageal carcinoma cases was (3.21 ± 2.33)%. However, the larger dose deviations in OARs (such as lens, parotid gland, optic nerve, and spinal cord) can be determined based on DVHs. The difference was particularly obvious for OARs with small volumes; for example, the maximum dose deviation for the lens reached (−6.12 ± 5.28)%. A comparison of the results obtained with Edose and Delta4 indicated that the Edose system could be applied for 3D pretreatment dose verification of IMRT. This system could also be utilized to evaluate the gamma passing rate of each treatment plan. Furthermore, the detailed dose distributions of PTVs and OARs could be indicated based on DVHs, providing additional reliable data for quality assurance in a clinic setting.

3 mm, and 2%/2 mm. Furthermore, the dose-volume histograms (DVHs) for planning target volumes (PTVs) as well as organs at risk (OARs) were analyzed using Edose. For dose verification of the 30 treatment plans, the average gamma passing rates of Edose reconstructions under the gamma criteria of 5%/3 mm, 3%/3 mm, and 2%/2 mm were (98.58 AE 0.93)%, (95.67 AE 1.97)%, and (83.13 AE 4.53)%, respectively, whereas the Delta4 measurement results were (99.14% AE 1.16)%, (95.81% AE 2.88)%, and (84.74% AE 7.00)%, respectively. The dose differences between Edose reconstructions and TPS calculations were within 3% for D 95% , D 98% , and D mean in each PTV, with the exception that the D 98% of the PTV-clinical target volume (CTV) in esophageal carcinoma cases was (3.21 AE 2.33)%. However, the larger dose deviations in OARs (such as lens, parotid gland, optic nerve, and spinal cord) can be determined based on DVHs. The difference was particularly obvious for OARs with small volumes; for example, the maximum dose deviation for the lens reached (À6.12 AE 5.28)%. A comparison of the results obtained with Edose and Delta4 indicated that the Edose system could be applied for 3D pretreatment dose verification of IMRT. This system could also be utilized to evaluate the gamma passing rate of each treatment plan. Furthermore, the detailed dose distributions of PTVs and OARs could be indicated based on DVHs, providing additional reliable data for quality assurance in a clinic setting. Because the Edose system constitutes a novel device using new technology for IMRT pretreatment dose verification, it was necessary to ensure its accuracy in clinical application. This study therefore aimed to evaluate the accuracy of the Edose system by comparing its reconstructed results with those measured using more established methods, namely the Delta4 device. [16][17][18] A total of 30 IMRT treatment plans are evaluated in the clinical application study. 2.B | The theory of 3D dose reconstruction in the Edose system The Edose system is a QA tool based on the patients' anatomy. This system uses the pixel values of images captured by EPID from treatment fields in air without a phantom/patient as input parameters.
The images are then reconstructed into a fluence map of the actually delivered beam through deconvolution and convolution. The CCCS algorithm was used for the 3D dose calculations and the 3D gamma evaluations. The theoretical formulas 19,20 of the fluence maps were reconstructed from the EPID image in the Edose system as follows: where where "a", which is proportional to monitor units (MU), is a coefficient denoting the EPID pixel values, "P ij " denotes the EPID image pixel values, and "K 1 (d ij )" is the EPID's scattering kernel and represents the energy scattering distribution after the interaction of the incident photon with EPID. In addition, l 1 and l 2 are the attenuation coefficients, which depend on the energy and the materials, c is the ratio constant, "K 2 (d ij )" is the fuzzy convolution kernel, which is the boundary factor for depicting the penumbra, l 3 affects the gradient of the penumbra, "ƒ(r ij )" is the Gaussian distribution function, which is introduced for shape correction, and e, r, and C r are obtained by comparing the profile after the reconstruction, mainly aimed at the large field to execute the adjustment. Furthermore, these three parameters reflect the upward curve of the saddle-shaped part of the profile. The variables "K 2 (d ij )" and "ƒ(r ij )" can correct the profile shape at various depths. c, l 1 , l 2 , l 3 , C r , e, and r are parameters for the detector kernel and were determined using the central point doses, which were measured in air for fields of 3 cm 9 3 cm to 25 cm 9 25 cm using an ionization chamber.
The parameters were constants once the fit procedure was com-

2.C | Clinical applications and evaluation of Edose
Prior to evaluation, comprehensive tests and evaluations to the Edose system were performed using the film and ionization chamber (IC) with uniform and human phantoms. The ionization chamber and radiochromic film were selected for measuring the point and planar doses for the single square and its combined fields and IMRT plans, and the corresponding results were compared to those reconstructed using Edose. The results showed that the point dose measured by the Edose agreed within 0.5% with the ionization chamber measurement in a uniform phantom. A minimum gamma pass rate of 95% was achieved for the comparison between Edose reconstructed dose maps and the planned dose maps when using the dose difference criterion of 5% of the maximum dose and a distance-to-agreement criterion of 3 mm (henceforth referred to as 5%/3 mm) and 3%/3 mm gamma criteria. The details are provided in our previously published manuscript. 21 In the present study, 3D dose reconstruction results obtained using the Edose system and measured with Delta4 were compared with those calculated by TPS to verify the feasibility of the Edose system. Thirty IMRT patient plans, including 10 nasopharyngeal carcinoma (NPC) plans, 10 esophageal carcinoma (EPC) plans, and 10 rectal cancer (REC) plans, were selected randomly for the 3D dose verification of Edose.

2.C.1 | Treatment plan
Thirty IMRT patient plans were optimized and calculated using an Data analysis was performed with Scandidos software (Scandidos, Uppsala, Sweden), which allows the user to compare the measured dose distribution for a complete treatment plan with the dose distribution predicted by the TPS. The ambient temperature was entered into the Delta4 software prior to every measurement because the diode detector response varied with temperature.

2.C.3 | Evaluation methods
An analytical method for 3D gamma evaluations 22 was applied to compare the 3D dose distributions that were reconstructed using HUANG ET AL.

| 99
Edose and measured using Delta4 with those obtained with the TPS.
A dose difference criterion of 5% of the maximum dose and a distance-to-agreement criterion of 3 mm were selected. A 3%/3 mm criterion and a more rigid 2%/2 mm criterion were also evaluated.
The gamma evaluation was global for both devices, and the threshold used for the dose analysis was more than 10% of the prescribed dose. The same areas that were used for the dose comparisons were defined as regions of interest (ROIs). A graphics processing unit (GPU) operation was employed in the system to improve the efficiency of the 3D gamma calculation. 23,24 Dose-volume histograms (DVHs) for both target areas and OARs were also compared between the TPS and the Edose system. The ΔD mean and V 40 analyses were performed in the bladder; and ΔD mean and V 20 analyses were performed in the femoral heads.

3.A | Gamma passing rate
The gamma passing rates obtained through the Edose reconstructions and Delta4 measurements were compared with the TPS calculations. Figure 1 shows the range of the gamma passing rates of 30 selected cases at the gamma criteria of 5%/3 mm, 3%/3 mm and 2%/2 mm. The narrow range showed the good reproducibility and stability of the response obtained using the Edose system. Overall, the range of the gamma passing rate detected with Delta4 was greater than that detected with Edose. Figure 2 illustrates the gamma passing rates for the REC, EPC, and NPC plans as well as all 30 plans under the three above-mentioned gamma criteria. The comparison results are expressed as the means AE SD in Fig. 2

3.B | DVH comparison between Edose and TPS
In comparison to the TPS calculations, the ΔD 95 , ΔD mean , and ΔD 98 of the PTVs obtained using Edose showed that the deviations were less than 3%, with the exception of the ΔD 98 found for the PTV-CTV in esophageal carcinoma patients, which was (3.21 AE 2.33)%, as shown in Table 2. The ΔD 95 indicates higher value of (À2.05 In other words, the values found for the deviation of these OARs were greater than AE 3%. Using the NPC plan as an example, Fig. 3 shows the 3D dose distributions in three axial slices, and Fig. 4 shows the DVH comparison of the PTVs and OARs obtained using the Edose and TPS. As shown in Fig. 3  The gamma passing rate reached 92% under the criteria of 5%/ 3 mm and 3%/3 mm. Currently, a gamma passing rate of at least 88% under 3%/3 mm is widely used to indicate the feasibility of IMRT plans, 25 and many radiotherapy central believe that the gamma evaluation method is reliable and effective for IMRT treatment verification. The results reflect in Fig. 2 show that Edose is more suitable for QA of IMRT due to its small deviation and high stability.  Fig. 4, was obtained because these OARs have a smaller volume; smaller dose differences would lead to a larger dose deviation in these OARs. 28 The same results were obtained in a previous study conducted by Chen. 29 The gamma passing rate is not sufficient for evaluating the feasibility of an IMRT plan, because it F I G . 3. Comparison of dose distributions in axial, coronal and sagittal views obtained using Edose (dotted lines) and calculated using the TPS (solid lines) for one of the NPC patients. The gamma passing rate using the 3%/3 mm global criterion was 96.8%. The green isodose line represents the 45-Gy distribution, the blue isodose line represents the 54-Gy distribution, the pink isodose line represents the 60-Gy distribution, the red isodose line represents the 66-Gy distribution, and the black isodose line represents the 68-Gy distribution. As shown in Table 2, the deviation between the Edose reconstruction and the TPS calculation was greater than 5%, and this value was obtained for the clinical target areas located close to the skin surface, such as the PTV-GTV-N-R in NPC. This phenomenon is related that the dose distribution is reconstructed using the fluence maps based on EPID. The dose of the built-up region mainly come from the local dose deposition and relatively small contribution from the distal area, which would lead to an inability of the 3D dose verification system (Edose) to amend the dose to the built-up region. In general, only a slight effect on the dose verification accuracy for the IMRT plan was found because the target areas were not too close to the built-up area. However, a certain effect on in vivo measurements could be obtained, because the EPID as a detector for photon rays in the Edose system showed a low-energy response. 32 Therefore, subsequent versions of Edose should correct the scatter in the built-up area to achieve better dose verification results.

| CONCLUSIONS
This study tested a preliminary clinical application of 30 IMRT plans, and the results showed that the EPID-based 3D dose verification system (Edose) was a simple and convenient QA tool for IMRT pretreatment dose verification. The gamma analysis included comparisons with the Delta4 system to validate the accuracy and reliability of the Edose system. The system could provide more clinical data and information through a single measurement. Furthermore, this system could allow a more intuitive and effective assessment for pretreatment plan dose verification. In future work, we would apply more verification tools to compare and verify the EPID-based system, and proceed more detailed verification and clinical applicability of this system.

ACKNOWLEDGMENTS
The authors are grateful to Raydose Medical Technology Co., Ltd for providing their EDose system and the technical support of computer.

CONF LICT OF I NTEREST
The authors have no conflicts of interest to disclose.