Evaluation of Concomitant Imaging Dose in 4D-CBCT Guided Thoracic Radiotherapy

Abstract

Introduction

This retrospective study evaluated the imaging dose of 4D-CBCT in patients treated with thoracic radiotherapy.

Methods

A geometric model of a 4D-CBCT system was built for Monte Carlo (MC) calculations. Percentage depth dose (PDD) and profiles of the system were measured and compared with MC calculations to verify the model. Eight lung cancer patients were retrospectively selected for 4D-CBCT imaging dose assessment. For each patient, the imaging dose was calculated using the CBCT log file, simulating CT and the MC model. Organ at risk (OAR) dose parameters from the imaging dose were calculated, and the effect of the imaging dose on OAR dose parameters when the imaging dose was added to the treatment dose was analyzed.

Results

Deviations between MC calculations and measurements for PDD and profiles were within 1.1% and 5%, respectively. For a pre-treatment 4D-CBCT scan, the mean lung and heart doses were 9.5 ± 1.4 mGy (mean value ± standard deviation, hereafter the same) and 9.7 ± 1.4 mGy, and the maximum spinal cord, esophagus and ribs doses were 17.6 ± 4.2 mGy, 17.1 ± 3.3 mGy and 37.8 ± 5.3 mGy, respectively. For an in-treatment 4D-CBCT scan, the dose parameters were 5.6 ± 1.8 mGy, 5.6 ± 1.5 mGy, 10.3 ± 3.5 mGy, 10.3 ± 4.3 mGy and 22.4 ± 5.3 mGy, respectively. The OAR dose parameters for in-treatment 4D-CBCT scans showed a relatively strong linear relationship with beam MU and number of frames. For patients receiving 5-fraction treatments and one pre-treatment and one in-treatment 4D-CBCT for each fraction, the increases in lung V20, lung V5, mean lung and heart doses, maximum spinal cord, esophagus and ribs doses were 0.04 ± 0.03%, 0.29 ± 0.23%, 76.3 ± 14.4 mGy, 76.5 ± 12.0 mGy, 76.3 ± 20.2 mGy, 70.5 ± 13.1 mGy and 101.2 ± 20.5 mGy, respectively. The increase in OAR dose parameters was proportional to the number of fractions.

Conclusion

The imaging dose of pre-treatment and in-treatment 4D-CBCT was calculated using a validated MC model. For 5-fraction treatments, the imaging dose will likely have minimal clinical impact on OAR dose parameters.

Keywords

4D-CBCT imaging dose Monte Carlo method thoracic radiotherapy in-treatment 4D-CBCT

Introduction

Cancer has been an increasingly serious threat to human health worldwide.¹ About half of patients are estimated to receive radiotherapy at some point after a diagnosis of cancer.² Image-guided radiation therapy (IGRT) has increasingly been recognized as a necessary part of precision radiotherapy. The flat-panel kV imaging device integrated into linear accelerator is the most commonly used in room image guidance method. It is capable of acquiring both 2D radiographs and 3D-CBCT images.³ Recently 4D-CBCT has been developed to address the significant respiratory motion artifacts in 3D-CBCT when imaging the thoracic region. Using 4D-CBCT, multiple phase-based CBCT datasets are generated which characterize the tumor motion at various phases of the breathing cycle. The respiration-correlated breathing signal for 4D-CBCT reconstruction is extracted from the CBCT projection data itself, eliminating the need for an external surrogate or implanted fiducial markers.⁴ 4D-CBCT has been progressively integrated into the lung cancer SBRT workflow.⁵

The imaging dose of 4D-CBCT is currently an issue concerning researchers and users.⁶ There are three reasons for the difference in imaging dose between 4D and 3D CBCT, which are explained here using the x-ray volume imaging (XVI) system (Elekta, Stockholm, Sweden) as an example. Firstly, more projections are required to obtain sufficient respiratory motion information, and therefore the acquisition time for 4D-CBCT (on the order of 3-4 min) is significantly longer than for 3D-CBCT (approximately 1 min).⁴ Secondly, the precision requirements of lung cancer SBRT treatment have resulted in institutions acquiring 2 to 3 4D-CBCT datasets for each fraction. The pre-treatment 4D-CBCT is used to measure and correct any misalignment of the time-weighted mean tumor position, while the in-treatment (or post-treatment) 4D-CBCT is used to assess any intrafraction patient movement. Some institutions use an additional pre-treatment 4D-CBCT scan to validate the correction applied.⁵ Thirdly, during in-treatment 4D-CBCT scan, the XVI flat-panel continues to acquire images at a constant rate of 5.5 frame/s, while the gantry rotation speed depends on the treatment plan and may change constantly. As a result, the projection angles may be uneven and the imaging dose may differ from the pre-treatment 4D-CBCT.

For the above reasons, the imaging dose of 4D-CBCT may be significantly greater than that of 3D-CBCT. Researchers have made efforts to reduce the imaging dose of 4D-CBCT by using fewer projections for reconstruction,⁶ but this is not yet clinically available. Therefore, understanding the magnitude of this dose is important in order to minimize its risk. Currently, imaging dose is not taken into account in radiotherapy treatment planning, which may lead to lower dose estimates for OARs.

Monte Carlo (MC) methods were the recommended dose calculation algorithms for kV imaging dose by AAPM TG report 180.³ Many researchers have evaluated the imaging dose of 3D-CBCT.⁷ Some previous studies have built models of the XVI system using MC methods^8-10 or treatment planning systems¹¹ and verified them with measurements in phantoms. Zhang et al evaluated the concomitant imaging dose and dose parameters to organs at risk (OARs) in 3D-CBCT guided thoracic radiotherapy using a Varian on-board imager (OBI), using MC methods.¹²

Fewer studies evaluated the imaging dose of 4D-CBCT. Nakamura et al evaluated the imaging dose and dose parameters to OARs from the 4D-CBCT module integrated in the Vero4DRT system (Mitsubishi Heavy Industries Ltd, Tokyo, Japan, and Brainlab AG, Feldkirchen, Germany).¹³ Yuasa et al used MC simulation to estimate the organ equivalent dose and effective imaging dose for 4D-CBCT performed with the OBI system, and evaluated the excess absolute risk of secondary cancer incidence.¹⁴ Thengumpallil et al analyzed the 4D-CBCT imaging dose of the XVI system using phantom measurements.¹⁵ To the best of our knowledge, no previous study has used the MC method to evaluate the imaging dose from 4D-CBCT performed with the XVI system, let alone from in-treatment 4D-CBCT scans.

The aim of this study is to evaluate the imaging dose of pre-treatment and in-treatment 4D-CBCT in patients treated with thoracic radiation therapy and its impact on OAR dose parameters based on MC calculations. First, geometric model of a CBCT system was built for MC calculations. The model was calibrated using absolute dose measurement and verified using percentage depth dose (PDD) and profiles measurements in water phantom. Subsequently, imaging doses from pre-treatment and in-treatment 4D-CBCT scans were calculated for 8 lung cancer patients, and the following were evaluated: (1) Dose distribution in patient CTs and dose parameters of typical OARs from pre-treatment and in-treatment 4D-CBCT scans. (2) The relationship between OAR dose parameters from in-treatment 4D-CBCT scans and beam MU or number of frames. (3) The increases in the OAR dose parameters when the imaging dose was added to the planned dose, compared with the OAR dose parameters of the planned dose alone.

Materials and Methods

Ethics Approval and Consent to Participate

The study was approved by the institutional review board of National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital in Shenzhen, China, on 28 February 2024 (approval number: JS2024-2-1). Participants provided written informed consent for agreeing the use of personal medical data prior to participating in the study. All patient details were de-identified in this study. The reporting of this study conforms to STROBE guidelines.¹⁶

Patient Data

Simulating 4DCT images and CBCT log files for 14 lung cancer patients previously treated in 2024 at our hospital were randomly selected for this retrospective study according to the approval of the local ethics committee. All patients were used for evaluating the correlation of number of kV frames acquired with in-treatment scan as a function of treatment beam monitor units (MUs). Eight of the patients were randomly selected for evaluating the concomitant imaging dose from 4D-CBCT scans (Table 1). Of the 8 patients, 4DCT images were acquired in 3 patients during free breathing (FB) and average intensity projection (AIP) CTs were used for MC calculations. For the other 5 patients, 3DCT images were acquired during deep inspiration breath hold (DIBH) achieved using the Active Breathing Coordinator (ABC) system (Elekta, Stockholm, Sweden) and were used for MC calculations. All patients were treated on an Infinity linear accelerator (Elekta, Stockholm, Sweden) with 2 or 3 volumetric modulated arc therapy (VMAT) beams with 6 MV flattening filter free (FFF) energy, arc length of 360° and MUs ranging from 505 to 2542 per beam. Since the in-treatment 4D-CBCT images were acquired during delivery of the first treatment beam as stated in our clinical protocol, the first treatment beams from each patient plans were used for evaluation in this study.

Table 1.

Basic Patient Information.

Patient No.	Breathing mode	Fractional dose (Gy)	Total dose (Gy)	No. of VMAT arcs	MU of the first VMAT arc	No. of frames for pre-treatment 4D-CBCT	No. of frames for in-treatment 4D-CBCT
1	FB	10	50	2	2542	1000	869
2	FB	10	50	3	1764	1000	596
3	FB	6.5	32.5	3	1295	1000	535
4	DIBH	7	35	2	1274	1000	635
5	DIBH	7	35	2	1476	1000	711
6	DIBH	10	50	2	630	1000	487
7	DIBH	3	54	2	929	1000	487
8	DIBH	2	50	2	505	1000	396

CBCT System and 4D-CBCT Feature

This study evaluated the XVI system version 5.0.7.1-b2 integrated with the Infinity linear accelerator platform. The XVI system consists of a kV x-ray source (70∼150 kVp) with an amorphous silicon/cesium iodide flat-panel detector. It is capable of imaging in either radiographic, fluoroscopic, 3D-CBCT or 4D-CBCT modes. Volumetric images can be acquired with three different field-of-views (FOVs) at the isocenter, namely small (S), medium (M), and large (L), each using different collimators. The two filter options are F0, which uses no filter, and F1, which uses a bow-tie filter. Both collimators and filters are inserted in front of the kV source and can be changed manually.

4D-CBCT scans for lung patients in our hospital are usually performed with S20 collimator and F1 filter. For both pre-treatment and in-treatment 4D-CBCT scans, the acquisition preset parameters were set to 120 kV, 20 mA and 16 ms per frame. These settings were used in the modelling of the CBCT system. A pre-treatment 4D-CBCT scan has an arc length of 360°. It takes approximately 3 min and acquires 1000 frames. For in-treatment 4D-CBCT scans, the imaging panel acquires images at a constant rate of 5.5 frames/s, and the gantry movement is determined by the treatment plan delivery. In the case of this study, all patients were treated with VMAT beams with an arc length of 360°, so the in-treatment 4D-CBCT scans all have an arc length of 360°. An additional AcquisitionInterval parameter was defined and set to 0.1°. During the in-treatment imaging process, when the gantry does not move fast enough and thus the angle between two frames is less than the AcquisitionInterval, a kV source off projection image would be taken instead to avoid redundant images. Table 2 shows details of the imaging protocols for pre-treatment and in-treatment 4D-CBCT.

Table 2.

Comparison of the Imaging Protocols Between Pre-Treatment and in-Treatment 4D-CBCT.

	Pre-treatment 4D-CBCT	In-treatment 4D-CBCT
Field-of-view (FOV) and filter	S20, F1	S20, F1
Imaging parameters	120 kV, 20 mA, 16 ms	120 kV, 20 mA, 16 ms
Scan time	3 min	Depending on the plan
Scanning arc length	360°	360°
Sampling frequency	5.5 frames/s	5.5 frames/s
Number of frames	1000	Depending on the plan

Monte Carlo Code and Simulation Configuration

The TOPAS MC code¹⁷ is an advanced and user-friendly extension to Geant4.^18-20 TOPAS version 3.7 with Geant4 version 10.06.p03 was used in this work. The physics module “g4em-standard_opt0” was used. The production threshold for all particles was taken as 10 keV following a previous work.⁹ Particle induced x-ray emission (PIXE) was activated in all processes of the simulations.

CBCT System Modeling

The modeling of the CBCT system, including the x-ray tube and the collimation system, followed previous studies.^8-10 A phase space file was acquired at the exit plane of the F1 filter with 2 × 10¹² incident electrons in the x-ray tube. This phase space file was used for further calculations in phantoms and patient CTs.

Percentage Depth Dose and Lateral Profiles Measurements

PDD and lateral profiles measurements in water were carried out using a computer-controlled scanning phantom Beamscan (PTW) with a waterproof PTW 31021 ionization chamber of 0.07 cm³ nominal active volume. Another PTW 31021 ionization chamber was used as field reference detector in all measurements. PDD and lateral profiles at 1 cm depth were acquired using the WaterTankScans software version 4.2 (PTW) in step mode. The step size varied from 0.2 cm to 1 cm for PDD measurement, and was 0.5 cm for lateral profile measurements.

These measurements were taken using the CBCT system in a stationary gantry position at 270°, with the kV source at 0° for vertical irradiation. The source to surface distance (SSD) was set to 100 cm. The kV beam was delivered 300 frames, with nominal 20 mA and 16 ms per frame.

Absolute Dose Calibration

Absolute dose measurement was carried out with the PTW 30013 Farmer chamber calibrated for absolute dose in kV x-ray according to IPEMB code of practice.²¹ The measurement was carried out at the water phantom surface and 30 cm SSD. The dose to water at the water phantom surface was measured using the following equation

D_{w, z = 0} = M N_{K} B_{w} {[{(\frac{{\bar{μ}}_{e n}}{ρ})}_{w / a i r}]}_{a i r},

where

D_{w, z = 0}

is the dose to water when the chamber center is positioned at the surface of the phantom material, M is the instrument reading obtained with a chamber in air corrected to the standard pressure and temperature, N_K is the chamber air kerma calibration factor, B_w is the back-scatter factor, and

[{({\bar{μ}}_{e n} / ρ)}_{w / a i r}]_{a i r}

is the mass energy absorption coefficient ratio, water to air, averaged over the photon spectrum in air. The N_K, B_w and

[{({\bar{μ}}_{e n} / ρ)}_{w / a i r}]_{a i r}

used in this work were 0.893, 1.343 and 1.042, respectively.

The same experimental conditions described above were reproduced with TOPAS MC code. A calibration factor F was then calculated using the following equation:

F = D_{e x p} / D_{M C},

where D_exp is the absolute dose measured and D_MC is the dose calculated by MC.

Validation of the CBCT Model Using Heterogeneous Phantoms

Two different combinations of water equivalent (RW3, PTW, Germany) slabs and tissue mimicking material slabs representing cortical bone (SB3) and lung tissue (LN-300) (SunNuclear, USA) were used as heterogeneous phantoms to validate the accuracy of dose calculations. The lung phantom consisted of 2 cm of RW3 slabs, a 6 cm LN-300 slab and 8 cm of RW3 slabs from top to bottom, and the measurement depths were 1 cm and 9 cm. The bone phantom consisted of 2 cm of RW3 slabs, a 1 cm SB3 slab and 8 cm of RW3 slabs from top to bottom, and the measurement depths were 1 cm and 4 cm. The irradiation setup of the phantoms was the same as for the PDD and profiles measurements. The kV beam was delivered 1000 frames, with nominal 20 mA and 16 ms per frame. Dose measurements were carried out using LD-V1 films (Ashland, USA), and each measurement was repeated for 3 times. The use and calibration of the film generally followed the methodology of a previous study.²² The phantoms were CT scanned and the MC calculated doses using the same setup were compared with the measurements.

Dose Calculation in Water Phantom and Patient CTs

A water cube with dimensions of (80 cm)³ was used for dose calculations in water. For PDD and lateral profile calculations, the voxel size was set to (0.5 cm)³. When calculating absolute dose in ionization chamber, a geometry model of the PTW 30013 ionization chamber was built in order to calculate the dose collected in the active volume. Detailed descriptions of the geometries were found in previous papers²³ and provided by the manufacturers.

For 4D-CBCT dose calculations in patient CTs, the gantry angle of each frame was extracted from the CBCT log file, transformed into the kV source angle (ie 90° was added), and then written into the MC code using a script. The kV source angles were nearly equidistant for pre-treatment scans and non-equidistant for in-treatment scans, as the latter depended on the treatment plan. The kV source angle of each frame was used as the angle of the phase space source. Using the time feature system of TOPAS, the dose in patient CTs was calculated per frame with a time-dependent kV source angle and then automatically summed to obtain the total 4D-CBCT imaging dose. The isocenter of the kV beam was set to the same value as that used for treatment planning. For in-treatment scans, only frames with the kV source on were used for dose calculations. The CT HU to material converter was implemented following the technique developed by Schneider.²⁴ The voxel size was set to (0.3 cm)³, the same as the treatment planning dose calculation grid size.

In our institution, 4D-CBCT is most commonly used in SBRT treatments with 5 fractions, and occasionally in treatments with 10 or more fractions. For each treatment fraction, a pre-treatment and an in-treatment 4D-CBCT scan are usually used. To account for the usually ignored impact of imaging dose on OAR parameters, the imaging dose of all the 4D-CBCT scans used throughout the treatment was added to the planning dose. The single-fraction imaging dose was multiplied by the number of fractions to obtain the total imaging dose for the multi-fraction treatment. The OAR parameters of the planning dose with imaging dose were then compared to those of the planning dose alone to determine the increase in OAR dose parameters induced by 4D-CBCT scans. The sum of the planning dose and the imaging dose and the evaluation of the OAR dose parameters, including mean lung and heart doses, maximum spinal cord, esophagus and ribs doses, lung V20 (percentage volume of the whole lung receiving ≥ 20Gy) and V5 (percentage volume of the whole lung receiving ≥ 5Gy), were performed using MIM version 7.3.2 (MIM Software Inc., Cleveland, OH, USA).

To validate the AIP-derived imaging dose calculation, we used a 4D dose calculation method and compared the results with those calculated using AIP CTs. For each kV frame, it is assumed that the x-ray source would irradiate the ten phases of the 4DCT image sets evenly. This is a reasonable assumption, given that the CBCT sampling frequency (5.5 frames per second) is much higher than the breathing frequency (12-20 cycles per minute). Under this assumption, the 4D dose was calculated as the average of the doses calculated on all the individual phases of the 4DCT simulation scan using the kV frames at different angles, without considering the time dependence of the delivery sequence. Deformable registrations were performed between each 4DCT phase and the AIP CT. Then, the dose calculated for each phase could be deformed to the AIP CT, and the average dose could be computed.

Statistical Analysis

All statistical analysis were performed with IBM SPSS V22 software (IBM Incorporate, Armonk, USA). Values are presented as the mean ± standard deviation or median and inter-quartile range. Linear regression was used to model the relationship between dependent and independent variables. We adopted either a two-tailed unpaired t-test (normal distribution) or the Wilcoxon signed-rank test (non-normal distribution). A p-value of less than 0.05 (*p < 0.05) was considered statistically significant.

Results

Validation of the 4D-CBCT System Model

The comparison between PDD measured in water and MC calculations were given in Figure 1. The data were normalized to the dose at 1 cm depth along the central axis of radiation. The deviations between them were less than 1.1%. Figure 2 shows the comparison for cross-line and in-line lateral profiles. For both cross-line and in-line data, the discrepancy between measurements and MC calculations was less than 2.6% in plateau areas, and less than 5% near the fall-off area of the curves.

Figure 1.

MC Calculated and Measured Percentage Depth Dose (PDD) Curves of the 4D-CBCT System Using S20 and F1. the top Panel Shows the Normalized PDD Curves, and the Bottom Panel Shows Their Percentage Differences.

Figure 2.

MC Calculated and Measured (a) Cross-Line and (b) In-Line Profiles of the 4D-CBCT System Using S20 and F1. The Top Panels Show the Normalized Profile Curves, and the Bottom Panels Shows Their Percentage Differences.

The measured and MC calculated dose in heterogeneous phantoms were displayed in Table 3. Before the kV beam penetrated the tissue mimicking materials (at a depth of 1 cm), the differences between measured and calculated doses were less than 1%. After passing through the tissue-mimicking materials, the deviations were less than 4%.

Table 3.

MC Calculated and Measured Imaging Dose of the CBCT System in Heterogeneous Phantoms.

Phantom	Measurement depth (cm)	Measured dose (mGy)	Calculated dose (mGy)	Difference (%)
Lung phantom	1	28.7	28.9 ± 0.1	0.7
Lung phantom	9	18.7	19.1 ± 0.2	2.1
Bone phantom	1	28.9	29.2 ± 0.2	1.0
Bone phantom	4	19.6	20.3 ± 0.3	3.6

Imaging Dose from pre-Treatment and in-Treatment 4D-CBCT Scans

Figure 3 shows the typical imaging dose distribution on the CT of a thoracic patient from a single pre-treatment and in-treatment 4D-CBCT scan. They show similar patterns, although the magnitude of the dose differs. Relatively higher doses were found in bone and near the body surface. The statistical distributions of the dose parameters of the typical thoracic OARs from a single pre-treatment or in-treatment 4D-CBCT scan in 8 lung cancer patients are shown in Figure 4. The mean value ± standard deviations of these OAR dose parameters are shown in Table 4. No significant differences were observed in any of these dose parameters between the different breathing modes.

Figure 3.

Dose Distribution on a Patient's CT from a Single (a) pre-Treatment and (b) in-Treatment 4D-CBCT Scan, Each Showing the Transverse, Sagittal and Coronal Planes from Left to Right. the red Patch Indicates the Treatment Target.

Figure 4.

Dose parameters of the OARs from a Single Pre-treatment or In-Treatment 4D-CBCT Scan.

Table 4.

OAR Dose Parameters from a Pre-Treatment or in-Treatment 4D-CBCT Scan (Mean Value ± Standard Deviation).

OAR dose parameters (mGy)	Pre-treatment 4D-CBCT				In-treatment 4D-CBCT
Breathing mode	FB	DIBH	Both	p-value	FB	DIBH	Both	p-value
Lungs mean dose	10.7 ± 0.5	8.8 ± 1.3	9.5 ± 1.4	0.057	7.0 ± 1.6	4.8 ± 1.5	5.6 ± 1.8	0.097
Heart mean dose	10.5 ± 1.1	9.2 ± 1.4	9.7 ± 1.4	0.213	6.7 ± 1.2	4.9 ± 1.3	5.6 ± 1.5	0.100
Spinal cord maximum dose	18.9 ± 0.4	16.8 ± 5.4	17.6 ± 4.2	0.558	12.4 ± 3.5	9.0 ± 3.1	10.3 ± 3.5	0.207
Esophagus maximum dose	18.9 ± 0.7	16.1 ± 3.9	17.1 ± 3.3	0.288	12.5 ± 3.6	9.1 ± 4.5	10.3 ± 4.3	0.309
Rib maximum dose	40.3 ± 5.9	36.3 ± 4.9	37.8 ± 5.3	0.343	25.5 ± 3.4	20.5 ± 5.6	22.4 ± 5.3	0.212

The Relationship Between Beam MU or Number of Frames and Imaging Dose from in-Treatment 4D-CBCT Scan

The relationship between beam MU and the number of frames acquired during the in-treatment 4D-CBCT scan for 14 lung cancer patients is shown in Figure 5, along with the linear regression line. There is a relatively significant linear relationship between them. The linear regression equation obtained using least squares method is

Number of in - treatment kV frames = 0.2238 \times Beam MU + 296.57,

(1)

and the correlation coefficient r² is 0.8908. As the imaging dose is directly related to the number of frames acquired during the 4D-CBCT scan, it should also have a linear relationship with the beam MU. Figure 6 shows the relationship of imaging dose parameters of the OARs from in-treatment scans as a function of beam MUs for 8 lung cancer patients. All the investigated imaging dose parameters of the OARs show a relatively significant linear relationship with the beam MU. The correlation coefficient r² for mean lung dose, mean heart dose, maximum spinal cord dose, maximum esophagus dose and maximum rib dose are 0.8575, 0.7839, 0.7074,0.7482 and 0.7405, respectively. Figure 7 shows the relationship of OAR dose parameters from in-treatment scans as a function of number of kV frames acquired. The correlation coefficient r² for mean lung dose, mean heart dose, maximum spinal cord dose, maximum esophagus dose and maximum rib dose are 0.9833, 0.9826, 0.9653,0.9749 and 0.9886, respectively. The average OAR dose parameters per frame or per MU from an in-treatment 4D-CBCT scan are displayed in Table 5.

Figure 5.

Relationship of Number of kV Frames Acquired with in-Treatment Scan as a Function of Treatment Beam MUs.

Figure 6.

Relationship of Imaging Dose Parameters of the OARs from In-Treatment Scans as a Function of Beam MUs.

Figure 7.

Relationship of Imaging Dose Parameters of the OARs from in-Treatment Scans as a Function of Number of kV Frames Acquired.

Table 5.

OAR Dose Parameters per Frame or Per MU from an in-Treatment 4D-CBCT Scan (Mean Value ± Standard Deviation).

OAR dose parameters from an in-treatment 4D-CBCT scan	Per frame (μGy/frame)	Per MU (μGy/MU)
Lungs mean dose	9.5 ± 1.4	4.8 ± 1.3
Heart mean dose	9.5 ± 1.4	4.9 ± 1.6
Spinal cord maximum dose	17.4 ± 3.9	8.8 ± 2.8
Esophagus maximum dose	17.0 ± 3.4	8.5 ± 2.0
Rib maximum dose	38.2 ± 3.9	19.8 ± 6.7

The Increase in OAR Dose Parameters from 4D-CBCT Scans

Figure 8 shows the increase in OAR dose parameters from 5 fractions of pre-treatment and in-treatment 4D-CBCT scans for 8 lung cancer patients when imaging doses were added to the planned doses. For 5, 10 and 25 fractional treatments, the increases in OAR dose parameters are shown in Table 6. For treatments with 5 fractions, the increases in lung V20, lung V5, mean lung dose, mean heart dose, maximum spinal cord dose, maximum esophagus dose and maximum rib dose are 0.04 ± 0.03%, 0.29 ± 0.23%, 76.3 ± 14.4 mGy, 76.5 ± 12.0 mGy, 76.3 ± 20.2 mGy, 70.5 ± 13.1 mGy and 101.2 ± 20.5 mGy, respectively. For treatments with 25 fractions, the increases are 0.21 ± 0.16%, 1.52 ± 1.17%, 380.2 ± 72.1 mGy, 381.6 ± 60.1 mGy, 381.8 ± 100.8 mGy, 352.5 ± 66.5 mGy and 507.4 ± 103.8 mGy, respectively.

Figure 8.

The Statistical Distributions of the Increases in OAR Dose Parameters from 5 Fractions of Pre-treatment and In-treatment 4D-CBCT Scans.

Table 6.

The Increases in OAR Dose Parameters When Comparing the Treatment Plan Dose Including and Excluding the Imaging Dose from a Pre-Treatment and an In-Treatment 4D-CBCT Scan Per Fraction.

The increases in OAR dose parameters	5-fraction treatments	10-fraction treatments	25-fraction treatments
Lungs V20 (%)	0.04 ± 0.03	0.08 ± 0.06	0.21 ± 0.16
Lungs V5 (%)	0.29 ± 0.23	0.59 ± 0.46	1.52 ± 1.17
Lungs mean dose (mGy)	76.3 ± 14.4	152.3 ± 28.8	380.2 ± 72.1
Heart mean dose (mGy)	76.5 ± 12.0	152.8 ± 24.0	381.6 ± 60.1
Spinal cord maximum dose (mGy)	76.3 ± 20.2	152.7 ± 40.5	381.8 ± 100.8
Esophagus maximum dose (mGy)	70.5 ± 13.1	140.9 ± 26.7	352.5 ± 66.5
Rib maximum dose (mGy)	101.2 ± 20.5	202.9 ± 41.5	507.4 ± 103.8

Figure 9 shows the increase in OAR dose parameters from pre-treatment and in-treatment 4D-CBCT scans per treatment as a function of the number of treatment fractions. All the investigated OARs show a significant linear relationship between the increase in OAR dose parameters and the number of treatment fractions.

Figure 9.

The Increase in OAR Dose Parameters from Pre-Treatment and In-Treatment 4D-CBCT Scans Per Treatment as a Function of the Number of Treatment Fractions.

Comparison Between 4D Dose Calculation and AIP CT Dose

Table 7 shows the OAR dose parameters from a pre-treatment 4D-CBCT scan for a patient treated with FB. Both doses calculated using the 4D dose calculation method and on AIP CT are displayed. All the dose parameters calculated using AIP CT differed by less than 0.8% from those calculated using 4D dose method.

Table 7.

OAR Dose Parameters from a Pre-Treatment 4D-CBCT Scan for a Patient Treated with FB.

OAR dose parameters	4D dose (mGy)	Dose in AIP CT (mGy)	Difference (%)
Lungs mean dose	10.23	10.21	0.20
Heart mean dose	9.31	9.36	−0.53
Spinal cord maximum dose	19.08	19.09	−0.10
Esophagus maximum dose	18.96	19.07	−0.58
Rib maximum dose	33.23	33.49	−0.78

Discussion

In this study, the imaging dose of pre-treatment and in-treatment 4D-CBCT in patients treated with thoracic radiation therapy and its impact on dose parameters of OARs was evaluated using a model of the XVI imaging system based on MC calculations.

The main reasons for the discrepancies between MC calculated and measured PDDs and profiles could be the slight difference in component dimensions and x-ray tube spectra between the XVI system described in the literature and those used in our institution for the manufacturing error, as well as the measurement errors. The deviations shown in this study are close to similar previous studies.^8,14 The discrepancies between the calculated doses in heterogeneous phantoms and the measurements were acceptable.

No significant differences were observed in any of the OAR dose parameters between FB and DIBH patients. This may suggest that the preliminary findings of this study could be applied regardless of the patient's breathing mode.

Preliminary results comparing 4D dose calculations with those calculated using AIP CTs showed that the two methods produce very similar results for OAR dose parameters.

The overall higher magnitude of the dose distribution and OAR dose parameters of pre-treatment 4D-CBCT compared to in-treatment 4D-CBCT can be attributed to the difference in the number of kV frames. The number of kV frames for pre-treatment 4D-CBCT was 1000, while the number for in-treatment 4D-CBCT scans varied between 396 and 869, as shown in Figure 5. The imaging dose can be roughly estimated as proportional to the number of frames.⁶ Relatively higher doses were found in bone and near the body surface, both consistent with previous studies for 3D-CBCT.^8,12

Yuasa et al reported the average lung and heart mean doses of 143 mGy and 160 mGy, respectively, for lung cancer patients from 4D-CBCT performed with OBI system.¹⁴ Nakamura et al reported median ipsilateral lung mean dose of approximately 30 mGy and heart mean dose of approximately 35 mGy for lung cancer patients from 4D-CBCT performed with Vero4DRT system.¹³ The average lung and heart mean doses presented in this study from pre-treatment 4D-CBCT performed with XVI system were 9.5 mGy and 9.7 mGy, respectively, which are much lower than those reported in previous studies. These discrepancies were mainly due to the differences in imaging systems and protocols. For reference, previous studies have shown that the 3D CBCT doses for OBI and XVI systems differ to a large extent.²⁵ The authors discussed a few reasons for the differences observed. Firstly, the x-ray tubes are located at opposite locations, which determines were the highest doses are deposited on the patient's surface in combination with the acquisition angles used. Secondly, even at similar kVp settings, the x-rays exhibited different levels of filtrations. With such different beams, it was observed that, at similar mAs, more dose is deposited using the OBI unit compared with the XVI unit. The authors didn’t conclude that how different collimator or filtration designs affect the imaging dose. Therefore, this study is only applicable to the XVI system and the scanning protocols used in this paper.

Due to differences in patient geometry, positioning and number of kV frames for in-treatment imaging, the OAR dose parameters showed large inter-patient variations. The variations are in the same range as previous similar studies.^13,14 In this study, only 8 patients were selected for MC calculations due to the high cost of computational resources. The clinical findings in this work have limited generalizability due to the small sample size. Future studies could include more patient cases using more computational resources or fast MC algorithms to reduce the variance of the results, and stratified analysis could be used by breathing mode and patient geometry (BMI and anatomical dimensions) to gain further insight. However, variation due to inherent differences between patients cannot be reduced by using more patient samples.

The number of kV frames is considered to be approximately proportional to the imaging dose and has been used for rough imaging dose assessment.^4,6 This relationship is also shown in Figure 7. The imaging dose from the in-treatment 4D-CBCT scan is related to the treatment plan and more specifically to the beam MU. For in-treatment 4D-CBCT, the beam MU is negatively correlated with the gantry rotation speed, thus positively correlated with the number of frames acquired (Figure 5), thus positively correlated with the dose parameters to OARs (Figure 6). This relationship can be used in further studies with more patient cases to develop a predictive model for in-treatment 4D-CBCT imaging dose using beam MU as an input. Although the linear relationship between imaging dose and number of frames is more obvious, a predictive model using beam MU allows prediction at the treatment planning stage, which may be more clinically meaningful. At this stage, a rough prediction can be made using the average OAR dose parameters per frame or per MU from an in-treatment 4D CBCT scan shown in Table 5.

Most previous studies have not evaluated the difference between OAR dose parameters with and without imaging dose because it was not considered clinically significant.^13,14 While imaging dose is not currently considered in radiotherapy treatment planning, this work provided predictions of the increase in OAR dose parameters in treatment planning when 4D-CBCT imaging dose is added, and demonstrated that the imaging dose may have a clinically relevant impact on OAR dose parameters when used extensively. For 4D-CBCT scans in 5-fraction treatments, the increase in OAR dose parameters were insignificant, as shown in Figure 8. There is a significant linear relationship between the increase in OAR dose parameters and the number of treatment fractions (Figure 9). Therefore, special care should be taken when using 4D-CBCT in conventional radiotherapy, which has more treatment fractions. For example, when using 4D-CBCT with 25 treatment fractions, which is unusual in clinical practice, there will be an apparent average increase of 1.52% in lung V5.

Due to the higher LET of kV radiation, there is an increase in relative biological effectiveness (RBE) of absorbed doses of radiation from kV CBCT sources in reference to megavoltage doses. Previous works have determined the RBE of kV CBCT sources using MC calculations, with results ranging from 1.1 to 1.5.²⁶ The differences were mainly due to different biological endpoints and beam characteristics, and to the best of our knowledge, no consensus has been reached on the use of the RBE of kV CBCT sources. Therefore, in this study, the RBE of the kV CBCT dose was considered to be the same as the megavoltage dose, following previous studies.^3,27

AAPM TG report 180 has provided recommendations for the management of image guidance doses delivered during radiotherapy.³ It is recommended that imaging dose be considered part of the total dose at the treatment planning stage if the dose from repeated imaging procedures is expected to exceed 5% of the prescribed target dose. In the case of our study, the image dose was well below 5%. However, due to the strong influence of the imaging dose on lung V5 described above, we recommend that the imaging dose should be evaluated when using 4D-CBCT in treatments with 10 or more fractions.

Conclusion

In this study, the imaging dose of pre-treatment and in-treatment 4D-CBCT in patients treated with thoracic radiation therapy and its impact on dose parameters of OARs were evaluated based on MC calculations. The imaging doses of pre-treatment and in-treatment 4D-CBCT showed a similar distribution, but the magnitudes of the doses were different. The investigated OAR dose parameters for in-treatment 4D-CBCT showed a relatively strong linear relationship with the beam MU and the number of kV frames. There is a significant linear relationship between the increase in OAR dose parameters and the number of treatment fractions. For 5-fraction treatments, the imaging dose will likely have minimal clinical impact on OAR dose parameters.

Footnotes

Abbreviations

ORCID iDs

Wu Jianan

Zhao Man

Ethical Considerations

This retrospective study was approved by the review board of National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital in Shenzhen, China, on 28 February 2024 (No. JS2024-2-1). Written informed consent has been obtained. We confirm that all methods were carried out in accordance with relevant guidelines and regulations.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Shenzhen Science and Technology Program (N0. JCYJ20240813151606009 and No. JCYJ20220530153803008), National Natural Science Foundation of China (No. 12105367), Hospital Research Project (No. E010322001), Shenzhen Key Medical Discipline Construction Fund (No. SZXK013), Sanming Project of Medicine in Shenzhen Municipality (No. SZSM202211030) and Shenzhen High-level Hospital Construction Fund.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Sung

Ferlay

Siegel

, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209-249. doi: https://doi.org/10.3322/caac.21660

Chandra

Keane

Voncken

FEM

Thomas

Jr . Contemporary radiotherapy: Present and future. Lancet. 2021;398(10295):171-184. doi: https://doi.org/10.1016/S0140-6736(21)00233-6

Ding

Alaei

Curran

, et al. Image guidance doses delivered during radiotherapy: Quantification, management, and reduction: Report of the AAPM Therapy Physics Committee Task Group 180. Med Phys. 2018 May;45(5):e84-e99. doi: https://doi.org/10.1002/mp.12824

Liang

Lack

Zhou

Liu

Grills

Yan

. Intrafraction 4D-cone beam CT acquired during volumetric arc radiotherapy delivery: KV parameter optimization and 4D motion accuracy for lung stereotactic body radiotherapy (SBRT) patients. J Appl Clin Med Phys. 2019 Dec;20(12):10-24. doi: https://doi.org/10.1002/acm2.12755

Sonke

Rossi

Wolthaus

van Herk

Damen

Belderbos

. Frameless stereotactic body radiotherapy for lung cancer using four-dimensional cone beam CT guidance. Int J Radiat Oncol Biol Phys. 2009 Jun 1;74(2):567-574. doi: https://doi.org/10.1016/j.ijrobp.2008.08.004

Lau

BKF

Reynolds

Keall

, et al. Reducing 4DCBCT imaging dose and time: Exploring the limits of adaptive acquisition and motion compensated reconstruction. Phys Med Biol. 2022 Mar 7;67(6):065002. doi: https://doi.org/10.1088/1361-6560/ac55a4

Alaei

Spezi

. Imaging dose from cone beam computed tomography in radiation therapy. Phys Med. 2015 Nov;31(7):647-658. doi: https://doi.org/10.1016/j.ejmp.2015.06.003

Downes

Jarvis

Radu

Kawrakow

Spezi

. Monte Carlo simulation and patient dosimetry for a kilovoltage cone-beam CT unit. Med Phys. 2009 Sep;36(9):4156-4167. doi: https://doi.org/10.1118/1.3196182

Spezi

Downes

Radu

Jarvis

. Monte Carlo simulation of an x-ray volume imaging cone beam CT unit. Med Phys. 2009 Jan;36(1):127-136. doi: https://doi.org/10.1118/1.3031113

10.

Thing

Bernchou

Mainegra-Hing

Hansen

Brink

. Hounsfield unit recovery in clinical cone beam CT images of the thorax acquired for image guided radiation therapy. Phys Med Biol. 2016 Aug 7;61(15):5781-5802. doi: https://doi.org/10.1088/0031-9155/61/15/5781

11.

Alaei

Spezi

Reynolds

. Dose calculation and treatment plan optimization including imaging dose from kilovoltage cone beam computed tomography. Acta Oncol (Madr). 2014;53(6):839-844. doi: https://doi.org/10.3109/0284186X.2013.875626

12.

Zhang

Chen

, et al. Concomitant imaging dose and cancer risk in image guided thoracic radiation therapy. Int J Radiat Oncol Biol Phys. 2015 Nov 1;93(3):523-531. doi: https://doi.org/10.1016/j.ijrobp.2015.06.034

13.

Nakamura

Ishihara

Matsuo

, et al. Quantification of the kV X-ray imaging dose during real-time tumor tracking and from three- and four-dimensional cone-beam computed tomography in lung cancer patients using a Monte Carlo simulation. J Radiat Res. 2018 Mar 1;59(2):173-181. doi: https://doi.org/10.1093/jrr/rrx098

14.

Yuasa

Shiinoki

Onizuka

Fujimoto

. Estimation of effective imaging dose and excess absolute risk of secondary cancer incidence for four-dimensional cone-beam computed tomography acquisition. J Appl Clin Med Phys. 2019 Nov;20(11):57-68. doi: https://doi.org/10.1002/acm2.12741

15.

Thengumpallil

Smith

Monnin

Bourhis

Bochud

Moeckli

. Difference in performance between 3D and 4D CBCT for lung imaging: A dose and image quality analysis. J Appl Clin Med Phys. 2016 Nov 8;17(6):97-106. doi: https://doi.org/10.1120/jacmp.v17i6.6459

16.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Ann Intern Med. 2007;147(8):573-577. doi: https://doi.org/10.7326/0003-4819-147-8-200710160-00010

17.

Perl

Shin

Schumann

Faddegon

Paganetti

. TOPAS: An innovative proton Monte Carlo platform for research and clinical applications. Med Phys. 2012;39(11):6818-6837. doi: https://doi.org/10.1118/1.4758060

18.

Agostinelli

Allison

Amako

, et al. GEANT4—a simulation toolkit. Nucl Instrum Methods Phys Res A. 2003;506(3):250-303. doi: 10.1016/S0168-9002(03)01368-8

19.

Allison

Amako

Apostolakis

, et al. Geant4 developments and applications. IEEE Trans Nucl Sci. 2006;53(1):270-278. doi: https://doi.org/10.1109/TNS.2006.869826

20.

Allison

Amako

Apostolakis

, et al. Recent developments in Geant4. Nucl Instrum Methods Phys Res A. 2016;835(1):186-225. doi: https://doi.org/10.1016/j.nima.2016.06.125

21.

Klevenhagen

Aukett

Harrison

Moretti

Nahum

Rosser

. The IPEMB code of practice for the determination of absorbed dose for x-rays below 300kV generating potential (0.035mm Al-4mm Cu HVL; 10-300kV generating potential). Institution of Physics and Engineering in Medicine and Biology. Phys Med Biol. 1996 Dec;41(12):2605-2625. doi: https://doi.org/10.1088/0031-9155/41/12/002

22.

Masella

Murphy

Bazalova-Carter

. Characterization of a new low-dose and low-energy Gafchromic film LD-V1. J Appl Clin Med Phys. 2024 Dec;25(12):e14531. doi: https://doi.org/10.1002/acm2.14531. Epub 2024 Sep 11. PMID: 39259853; PMCID: PMC11633805.

23.

Gomà

Sterpin

. Monte Carlo calculation of beam quality correction factors in proton beams using PENH. Phys Med Biol. 2019 Sep 17;64(18):185009. doi: https://doi.org/10.1088/1361-6560/ab3b94

24.

Schneider

Bortfeld

Schlegel

. Correlation between CT numbers and tissue parameters needed for Monte Carlo simulations of clinical dose distributions. Phys Med Biol. 2000 Feb;45(2):459-478. doi: https://doi.org/10.1088/0031-9155/45/2/314

25.

Song

Kamath

Ozawa

, et al. A dose comparison study between XVI and OBI CBCT systems. Med Phys. 2008 Feb;35(2):480-486. doi: https://doi.org/10.1118/1.2825619

26.

Kirkby

Ghasroddashti

Poirier

Tambasco

Stewart

. RBE Of kV CBCT radiation determined by Monte Carlo DNA damage simulations. Phys Med Biol. 2013 Aug 21;58(16):5693-5704. doi: https://doi.org/10.1088/0031-9155/58/16/5693

27.

Alaei

Ding

Guan

. Inclusion of the dose from kilovoltage cone beam CT in the radiation therapy treatment plans. Med Phys. 2010 Jan;37(1):244-248. doi: 10.1118/1.3271582