Investigations of organ and effective doses of abdominal cone-beam computed tomography during transarterial chemoembolization using Monte Carlo simulation

C-arm cone-beam computed tomography (CBCT), performed using an angiographic system that rotates a C-arm-mounted, flat-panel detector around the patient, is an imaging technique capable of yielding three-dimensional (3D) volumetric images. Although CBCT is useful for providing additional information on anatomical relationships, detecting tumors, determining the feeding arteries of tumors, and identifying the distribution of the contrast agent injected through the catheter [1], the additional radiation dose resulting from extra 3D imaging acquisitions during interventional procedures is difficult to evaluate in a patient [2].

The effective dose (ED) is considered the most appropriate quantity for estimating the stochastic risk of exposure to ionizing radiation. However, the complexity of dose calculations for C-arm CBCT complicates performing total ED estimations for patients who have undergone interventions, including fluoroscopic procedures and C-arm CBCT runs. Some studies have investigated the patient dose for abdominal CBCT procedures based on Monte Carlo simulations or Thermoluminescent Dosimeter (TLD) measurements [2‐6]. To more rapidly estimate the ED, suitable conversion factors should be applied to the dose–area product (DAP) values [7]. Suzuki et al. surveyed three types of angiographic systems from three manufacturers and used three sizes of human-shaped phantoms with Monte Carlo simulations and TLD measurements to assess the doses and effects of the phantom size on the EDs for abdominal C-arm CBCT procedures [2, 5]. Their benchmark studies provided an important reference for abdominal CBCT dose investigations, and demonstrated that conversion factors are protocol specific and may differ among angiographic systems [2, 5].

This study estimated the organ dose, ED, and conversion factors for abdominal C-arm CBCT during transarterial chemoembolization (TACE) by using Monte Carlo simulations for the angiographic systems not included in previous studies. The C-arm rotation angle effects on the organ dose based on the simulations were also investigated. Additionally, the relationship between the ED and patient body mass index (BMI) was investigated, and conversion factors that convert the DAP to ED based on the BMI categories for C-arm CBCT acquisitions in TACE were proposed for more detailed ED estimations.

TACE is an angiographic procedure used to treat patients with hepatic tumors by injecting chemotherapeutic drugs into the selected hepatic artery [13]. To achieve more detailed patient dose investigations during these procedures, the ED obtained through fluoroscopy and CBCT imaging acquisitions should be individually evaluated. Multiple vendors have offered C-arm CBCT applications for angiographic systems, and the dose performance may vary because of the effects of varying designs on the beam quality as well as different protocol settings. To provide additional dose assessment information on abdominal CBCT, in addition to using popular angiographic units used in previous studies [2‐6], organ dose and ED for C-arm CBCT acquisitions during hepatic TACE with the angiographic system without AEC capability when performing CBCT acquisitions by using Monte Carlo simulations were investigated.

Studies have investigated the doses for abdominal CBCT, as summarized in Table 3. The mean ED for the CBCT run was determined as 3.5 ± 0.5 mSv in this study. These values were slightly higher than the dose calculated according to the same Monte Carlo technique for the medium phantom with C-arm CBCT acquisitions with a GE INNOVA 4100 (3.1 mSv) but lower than the dose calculated for the large phantom with C-arm CBCT acquisitions using the same GE angiographic system (3.8 mSv) in the study by Suzuki [2].

In this study, BMI scores were used to analyze the relationship between patient size and the ED. The results revealed that the ED during C-arm CBCT acquisitions decreased with increased patient size, which is consistent with the findings of Wielandts et al. [9, 14] but contrasted with those of Ector et al. [8] and Suzuki et al. [2, 5]. Suzuki et al. demonstrated that the DAPs and ED increased with increased phantom size in the abdominal CBCT procedures with all three angiographic systems they investigated [2]. Wielandts et al. calculated CBCT doses for the ablation of arrhythmias and reported that ED is inversely related to patient BMI; this tendency may be because of the very limited increase in the DAP with the BMI despite AEC execution [14]. In this study, because of the technical specifications of the investigated angiographic system, the exposure parameters were preset in the CBCT procedures, regardless of patient size variations, and AEC was not activated in the acquisitions. Thus, DAP values deviated only slightly among the patients in different BMI groups. When the same exposure is used for patients with a higher BMI, the X-ray beam is attenuated more before being absorbed by the organ; thus, the organ doses and EDs would be lower for patients with a higher BMI than for those with a lower BMI. The effects of patient size on the ED for CBCT in systems with fixed-exposure techniques would be similar to those in systems with limited AEC modulations.

The establishment of conversion factors provide an approach for ED estimations when the DAP is available during angiography; however, the conversion factors differ among the angiographic systems and are specific to the used imaging protocols. Suzuki et al. estimated the ED to DAP ratios for three phantom sizes by using three types of angiographic systems for 3D abdominal imaging procedures [2]. Notably, in their survey, patient height and weight affected the ED to DAP ratios slightly, whereas the ED to DAP ratios were 0.37–0.45, 0.26–0.32, and 0.13–0.15 mSv·Gy^− 1·cm^− 2 on the Philips Allura Xper FD20/10, GE INNOVA 4100, and Siemens AXIOM Artis dTA systems, respectively, and the conversion factors were estimated to be approximately 0.4, 0.3, and 0.15 mSv·Gy^− 1·cm^− 2 on these three systems. Thus, Suzuki et al. concluded that the ED for each patient can be easily estimated using a suitable conversion factor set for each angiographic system [2].

To more conveniently evaluate the EDs for patients, our methodology was based on Monte Carlo simulations, and patient data were collected as input for dose calculations. In this study, the mean ED to DAP ratio was 0.27 ± 0.04 mSv·Gy^− 1·cm^− 2 on a Shimadzu BRANSIST safireVC17 for all patients, and the ratios decreased with increased BMI. The trend was similar with the results of Suzuki et al., and studies have reported that the body volume percentage in the exposure field decreases with increasing phantom size, thus contributing to the effects [2, 5]. The mean ED to DAP ratio estimated from all patients can serve as a conversion factor for easier ED estimations when the DAP is available. However, because of the strong effects of patient BMI on the ED to DAP ratios, using conversion factors without considering patient size may result in the overestimation of ED for obese patients. Using conversion factors adapted to patient size in addition to the DAP can serve as a feedback mechanism for providing clinicians with more details on ED estimations when performing C-arm CBCT.

The variation in the relative position of the organs to the X-ray tube during the C-arm rotation explains the fluctuations in the projection-by-projection radiation dose. The dose was higher when the organ was irradiated by the incident beam. During the C-arm rotation, dorsal organs, such as kidneys, received a higher dose during the major portions of the projections because the C-arm rotated from the left to the right side of the patient and went through the posterior side during the acquisitions. Among the 44 projections, dose variation was the highest for the stomach, followed by the liver, possibly because the stomach and liver were irradiated directly in the start and end positions, respectively, of the C-arm rotation. The dose absorbed by the liver increased during the rotation because the liver was more directly irradiated with the incident beam aimed at the anterior right side of the patient. By contrast, the stomach absorbed higher doses in the initial projections because of the direct X-ray irradiation in the beginning of the rotation trajectory, and the dose decreased during the rotation.

When totaling the organ doses from all 44 projections, the simulated results revealed that total doses to the organs in the upper abdomen were higher than those to the other organs. This is because the liver was the target organ during C-arm CBCT acquisitions. Therefore, the radiation window was always positioned in the upper abdomen during the C-arm rotation. In this study, dorsal organs, such as the kidney and adrenal gland, received the highest total dose during C-arm CBCT acquisitions, which likely occurred because the C-arm was rotated through the posterior side of the patients during the acquisitions, and the mentioned organs were localized in the direct-irradiated FOV for major portions of the projections during the exposure procedure. Notably, this phenomenon is in strong concordance with the results of Suzuki et al. [2, 5].

ED is calculated as a weighted sum of organ doses; therefore, its value is mainly determined by the organs that are highly irradiated and those that have more crucial weighting factors [9]. Based on the finding that the organ dose would vary in all radiation projections, ED for patients can be varied by adjusting the C-arm CBCT rotational angle. The stomach is one of the most radiosensitive organs (tissue weight factor, 0.12) [10] in the irradiated FOV; therefore, decreasing stomach dose by adjusting the C-arm CBCT rotational angle may lead to largely decrease total ED. C-arm rotation angle, FOV locations as well as the C-arm rotating around the anterior or posterior sides of the patients markedly affected patient doses, and this indicated that dose reduction strategies can be further manipulated from C-arm rotation angle setting or X-ray irradiated field location.

Our study has some limitations. First, the applicability of the results may be restricted. A system without AEC for C-arm CBCT applications was used in this study, and the results may be applicable to similar system configurations but not to those with AECs. However, the methodology described herein can still be used as a reference for patient dose evaluations of C-arm CBCT acquisitions. To provide more feasible clinical applications, the conversion factors for C-arm CBCT acquisitions should be further evaluated for different protocols and other angiographic systems from diverse manufacturers. Second, the use of the BMI as a patient size indicator has limitations. For example, a muscular patient with a narrow waist and an overweight patient can have a similar BMI, and a patient with ascites may have a low BMI but increased abdominal girth, which would affect the organ dose as well as the ED; this may not be reflected in the simulations.

We calculated the organ dose and ED according to a Monte Carlo technique for C-arm CBCT acquisitions during TACE by using a Shimadzu BRANSIST safireVC17 system. The ED to DAP ratios may differ with the protocols, systems, and patient sizes; however, overall, both ED and ED to DAP ratios decrease with increasing patient size. Suitable conversion factors for C-arm CBCT acquisitions facilitate the use of DAPs for estimating the ED during CBCT procedures and thus provide convenient patient dose estimations. The radiation dose absorbed by patients can be varied by adjusting the C-arm CBCT rotational angle settings, and dose reduction strategies can be further manipulated.