Comparing organ and effective dose of various CT localizer acquisition strategies: A Monte Carlo study

Abstract

Background

CT examinations commonly start with the acquisition of one or two localizer radiographs (2D localizers). Recently, a manufacturer introduced the option to perform a heavily filtrated low‐dose helical scan as a localizer acquisition. To compare the dose of one or two 2D localizer acquisitions to the dose of a 3D localizer acquisition, one cannot simply compare the CTDIs of the different acquisition techniques, because of the use of different geometries and spectra.

Purpose

To compare the organ and effective dose for various CT localizer acquisition techniques.

Methods

A Geant4‐based Monte Carlo simulation, replicating a clinical wide‐area CT scanner was developed and validated. Various localizer acquisition strategies were simulated: Anterior‐posterior (AP) alone, PA alone, combined AP+lateral (LAT), and PA+LAT 2D localizers, and an Ag‐filtered 3D localizer acquisition. Validation was performed by measuring and simulating CTDI₁₀₀ in both the periphery and the center of a CTDI phantom. The software was subsequently used to estimate organ and effective doses for localizers for chest, abdomen + pelvis, and the combined chest, abdomen, and pelvis exams. As representations of patients, eight ICRP computational phantoms (adult, 15‐, 10‐, and 5‐year, both male and female) and five female and five male XCAT phantoms with various BMIs were used. The dose of the various strategies was compared to the current clinically‐implemented AP+LAT localizers.

Results

CTDI₁₀₀‐measurements and simulations within the CTDI‐phantom differs by a maximum of 8.1% and by an average of 0.9%. For chest, the average effective doses for AP, PA, AP+LAT, and the 3D localizer are 0.10, 0.07, 0.32, and 0.22 mSv, respectively. The organ dose to the breast varies the most across the various localizer strategies and is, on average, 0.17, 0.03, 0.44, and 0.33 mGy, in the same order. For abdomen, the average effective doses are 0.11, 0.07, 0.36, and 0.25 mSv for the AP, PA, AP+LAT and the 3D localizer, respectively. The organ dose to the stomach varies the most across the various localizers and is on average 0.14, 0.08, 0.58, and 0.30 mGy, in the same order. The PA‐only localizer results in the lowest organ dose to the most radiosensitive organs and the lowest effective dose. For the chest exam, compared to AP+LAT, the PA+LAT results in a 7 ± 2% effective dose reduction (mean ± standard deviation), while the 3D localizer results in a 21 ± 3% effective dose reduction. Using AP or PA only would result in 69 ± 2% and 76 ± 2% reduction, respectively. For the abdomen exam, also compared to AP+LAT, PA+LAT results in 6 ± 2% effective dose reduction, while the 3D localizer results in a 20 ± 5% reduction. Using AP or PA only would result in 69 ± 5% and 76 ± 4% reduction, respectively.

Conclusions

Using a PA localizer results in a lower or equivalent organ dose in the most radiosensitive organs, and a lower effective dose compared to an AP localizer for both chest and abdomen+pelvis exams. Compared to a two‐localizer strategy, the 3D localizer results in a lower effective dose in both the chest and abdomen+pelvis region.

1. INTRODUCTION

Prior to the acquisition of the actual CT image, it is common to first acquire one or two localizer radiographies (2D localizers). The information resulting from these scans is used for automatic tube current modulation used during the CT acquisition and to set the correct region to be scanned. Some CT systems, by default, use only one localizer, mostly in the anterior‐posterior direction (AP). Others use two acquisitions; AP and a lateral (LAT) direction. Recently, one manufacturer introduced an option to replace the 2D localizers with a single low‐dose helical acquisition (3D localizer). This 3D localizer acquisition uses a heavily filtered beam, in which the default 0.4 mm copper filter is replaced by a 0.5 mm silver filter. As a result of the 3D localizer acquisition, the CT system shows the axial planes of the reconstructed volume, a synthetic AP/PA localizer image, and a synthetic LAT localizer image. Based on the selected protocol and the information from the 3D localizer, the CT system can automatically determine the correct start and end positions of the final CT acquisition.

The dose used for these 2D localizers may vary for various CT manufacturers and for the various body regions (e.g., head, trunk, extremities). It is generally assumed that the absorbed dose of the localizers is negligible compared to the actual CT exam. However, especially for low‐dose protocols, this is debatable. Previous estimations of localizer doses, performed for Siemens CTs, showed an effective dose in the range of 0.08 to 0.18 mSv for the chest. ¹ , ² According to the Lung Cancer Screening CT Protocols of the American Association of Physicists in Medicine (AAPM), the resulting effective dose will be less than 1 mSv for a standard‐sized patient. ³ So, depending on the localizer acquisition strategy, the effective dose of the localizer can be as much as 20% of the effective dose of the final acquisition for lung cancer screening CT. Dose comparisons for other systems, and especially for a 3D localizer using additional filtration, have not been previously performed. Therefore, it is unclear if the same dose levels are used for these new localizer protocols. To compare the dose of one or two 2D localizers between each other or to the dose of a 3D localizer acquisition, one cannot simply compare the CTDIs of the two acquisition strategies. First, the dose distribution in the body will be different, mainly because the direction of the x‐ray beam differs, but also because the spectra may differ. Secondly, for comparing the 3D localizer to the commonly used 2D localizer, the CTDI_vol for a 2D localizer is defined differently from that of a rotational scan. The CTDI_vol for a 2D localizer (called scan projection radiograph; SPR by IEC) is defined in IEC 60601‐2‐44:2009 as:

$$CTD{I}_{vol,spr}=\frac{CTD{I}_w}{Q{t}_r}·\frac{N·T}{V_{table}}·Q_{spr}$$ (1)

where CTDI_w is measured using the same collimation and x‐ray spectrum as during the 2D localizer (SPR) acquisition but using a sequential rotating acquisition. Qt _r is the x‐ray tube current time product used for this measurement. N·T is the collimation for the 2D localizer. V_table and Q_spr are the table speed and tube current during the 2D localizer acquisition.

Because comparing the CTDI values directly does not result in the correct comparison of patient‐absorbed dose nor effective dose, the goal of this research is to estimate these values for a range of patient characteristics for various 2D localizers and the 3D localizer using Monte Carlo simulations.

2. MATERIALS AND METHODS

For the dose evaluation of the 2D localizers and the 3D localizers, a clinical 320‐row CT system (Aquilion One PRISM Edition, Canon Medical Systems, Otawara, Japan) was used for the measurements and simulations in Monte Carlo software.

2.1. CT scanner and acquisition technique

The 320‐row CT system is equipped with 0.5 × 0.5 mm² effective detector element sizes. The focal spot to iso‐center distance is 60 cm and has a maximum field of view (FOV) of 500 mm. The 2D localizer is acquired by using a 2 mm collimation, a FOV of 500 mm, and a tube voltage of 120 kV. In addition to the inherent filtration and a bowtie filter, a Cu filter with a nominal thickness of 0.4 mm is used. The default implementation is that the system uses combined AP and LAT localizers. For the 3D localizer, a 4 cm collimation and a pitch of 1.381 is used, while the FOV, tube voltage, and bowtie filter are the same as for the 2D localizer acquisition. However, the Cu filter is replaced by an Ag filter with a nominal thickness of 0.5 mm. The CTDI_vol for the 2D localizer given by the system is the CTDI_vol,spr as defined by the IEC (Equation (1)). In this study, the 2D and 3D localizer acquisitions of three body parts were examined: chest alone, abdomen and pelvis, and the combined chest, abdomen, and pelvis. For the AP/PA and LAT localizers, the system uses a tube current of 30 and 100 mA, respectively, giving an indicated CTDI_vol,spr by the CT system of 0.10 and 0.33 mGy, respectively. For the 3D localizer it uses a tube current of 50 mA; indicating a CTDI_vol of 0.24 mGy.

2.2. Monte Carlo simulations

For the Monte Carlo simulations, a program was implemented using the Geant4 toolkit (version 10.7.2), ⁴ based on the Geant4 version of the simulation developed for the AAPM TG Report 195. ⁵ The standard EM physics (option 4; G4EmStandardPhysics_option4) was used and for each simulation 100 million photons were simulated.

2.2.1. Modeling the x‐ray spectrum, bowtie filter, and patient table

The x‐ray spectra, before the bowtie filter but including the Cu or Ag filtration, were modeled. For this, the dose at the iso‐center without and with additional filtration of various thicknesses of high‐purity copper (0.000, 0.257, 0.551, 0.808, 1.010, and 1.818 mm) were measured using a calibrated 0.6 cm³ ionization chamber (10 × 6–0.6 CT, Radcal, Monrovia, USA). The measured relative dose of the additionally filtered spectra was compared to those of the modeled spectra using an open‐source application for modeling x‐ray spectra (Spekpy 2.0.8). ⁶ The thickness of the Cu or Ag filtration of the modeled spectra were iteratively changed to minimize the summed absolute difference between the measured and modeled relative dose values. The thickness of the Cu and Ag filtration that resulted in the minimal summed absolute difference was used for the final simulations.

The shape of the bowtie filter was physically measured. The bowtie filter was not added as an object in the simulations but was simulated by applying a weighting factor to the emitted photons equal to $e^{-{\mu }_{Al}(E)d_{Al}(\phi )}$, where µ_Al(E) is the monochromatic attenuation coefficient depending on the energy of the emitted photon in aluminum and $d(\phi )$ the pathlength of the x‐ray through the bowtie when emitted at angle ϕ. ² While this method speeds up the simulations, it does not include the possibility of photons scattered in the bowtie filter and eventually reaching the body of the patient. However, given the distance from the bowtie filter to the patient's body, it is assumed that this effect can be neglected. To simulate the heel effect, a similar procedure was used. The extent of the heel effect was measured along the longitudinal direction, laterally centered using the same 0.6 cm³ chamber. This effect was translated to an effective path length in tungsten. During the simulations, the generated photons were given an additional weighting factor equal to $e^{-{\mu }_w(E)d_w(\theta )}$, where µ_w (E) is the attenuation coefficient for the specific photon energy in tungsten and d_w is the path length through the anode when emitted at angle θ.

The shape of the patient table was obtained by segmenting it from a CT acquisition. The material was modeled as carbon fiber with a density of 1.6 g/cm³.

To quantify the x‐ray output of the CT tube, CTDI₁₀₀ free‐in‐air was measured for both filtrations in the isocenter.

2.2.2. Validation of the model

The simulated shape of the bowtie filter was validated by measuring the relative dose compared to the dose at the isocenter at positions every 2 cm shifted laterally from the isocenter out to a maximum of 24 cm, with the tube fixed at the top position. These measurements were compared to the simulated dose distribution. The measured transmission of the table at the isocenter was compared to obtained with the simulations. For both comparisons, a 120 kV Cu filtered spectrum was used, and, the same 0.6 cm³ ionization chamber was used for the measurements.

To further validate the model, CTDI₁₀₀ measurements were both performed experimentally and in simulation. This was done by acquiring a scan with the same parameters as for the 3D localizer, except using a single axial acquisition instead of a helical one. CTDI₁₀₀ measurements were performed in the four periphery positions and in the center of the 32 cm diameter CTDI phantom. These were repeated using the Cu filter. The CTDI values at all five positions for both filters were compared to the simulated results. For both settings, a tube current of 400 mA and exposure time of 1 s was used.

2.2.3. Dose estimations in human models

After validation of the accuracy of the Monte Carlo simulations, the same program was used to simulate the PA, AP, and LAT localizers, as well as the 3D localizer. Two types of human computational patient models were used. The first type consisted of the ICRP computational models, of which the adult, 15, 10, and 5 year‐old phantoms were used, both female and male. ⁷ , ⁸ In addition, 10 XCAT phantoms were voxelized into 1.0 mm isotropic resolution. ⁹ Five of these phantoms were male and had a BMI ranging from 20 to 38 kg/m². The five female phantoms had a BMI ranging from 18 to 39 kg/m². Although the location of the organs is defined in the XCAT phantoms, their density and composition are not. Therefore, data from the ICRU publication 46 ¹⁰ was used to define the density and composition of each organ. An overview of the ICRP and XCAT phantom characteristics is given in Table 1. Next, for both phantom types, the classified voxels were assigned to the corresponding tissues listed in table 3 of the ICRP Report 103, if applicable. For the effective dose calculations, all tissues listed in table 3 of the ICRP Report 103 were considered, and the recommended tissue weighting factors were used. In the XCAT phantom, the red bone marrow is specifically classified as such, but this is not the case in the ICRP phantom. Therefore, to calculate the dose to the red bone marrow separately, the mass‐energy absorption coefficient method was used. ¹¹ For each phantom the body parted scanned was determined based on the CT localizer range as usually set what we would set as the localizer range clinically. Examples of the simulated scanned body parts are given in Figure 1 and the scan lengths for each phantom are given in Table 1.

TABLE 1.

Characteristics (gender, height, mass, and BMI) of the computational phantoms used and localizer scan lengths.

						Scan length [cm]
Phantom		Gender	Height [m]	Mass [kg]	BMI [kg/m2]	Chest	Abdomen	Chest/abdomen /pelvis
ICRP	Adult	Male	1.76	73.0	23.6	37	66	82
		Female	1.63	60.0	22.6	35	55	73
	15 years	Male	1.67	56	20.1	36	50	68
		Female	1.61	53	20.4	35	48	65
	10 years	Male	1.38	32	16.8	30	40	58
		Female	1.38	32	16.8	30	40	58
	5 years	Male	1.09	19	16.0	23	33	44
		Female	1.09	19	16.0	23	33	44
XCAT	PT86	Female	1.69	52.0	18.2	36	50	73
	PT170	Female	1.61	60.9	23.6	34	44	64
	PT76	Female	1.56	46.9	28.6	36	47	66
	PT182	Female	1.61	87.0	33.6	36	52	73
	PT147	Female	1.65	106	38.8	34	53	70
	PT150	Male	1.76	62.0	20.0	42	53	75
	PT118	Male	1.70	72.1	25.0	38	55	78
	PT164	Male	1.79	89.9	28.2	41	55	77
	PT144	Male	1.73	107	35.8	36	59	76
	PT184	Male	1.78	120	38.0	39	56	77

FIGURE 1.

Examples of the scan range used for the simulation of a chest localizer acquisition (red) and abdomen‐pelvis localizer (blue) for the 5‐year ICRP phantom (left) and the XCAT PT184 phantom. The chest‐abdomen‐pelvis localizer covers both scan ranges.

For each localizer acquisition technique, the organ and effective dose are calculated per dose‐length‐product (DLP). The DLP for the 2D localizer (DLP_spr ) can be derived by multiplying (1) with the scan‐length l:

$$DL{P}_{spr}=\frac{CTD{I}_w}{Q{t}_r}·\frac{N·T}{V_{table}}·Q_{spr}·l$$ (2)

In this equation the term $\frac{Q_{spr}·l}{V_{table}}$ is equal to the total tube current time product used for the 2D localizer acquisition.

The resulting simulated organ and effective doses are presented per acquisition. For the acquisitions, the default tube current settings for the CT system are simulated: 30 mA for the AP and PA localizer, 100 mA for the LAT localizer, and 50 mA for the 3D localizer. The dose values of localizer strategies using PA only, AP only, AP+LAT, PA+LAT, and a 3D localizer are compared for each scanned body region.

3. RESULTS

3.1. Modeling the x‐ray spectrum

A Cu filter thickness of 0.36 mm and an Ag filter thickness of 0.55 mm minimized the difference between measured and modeled relative dose values. These thicknesses both deviate 10% from their nominal values of 0.4 and 0.5 mm, respectively. Reasons for these deviations may be manufacturing tolerances, inaccuracies in the measurements or the modeling, and the deviation of the assumed aluminum filtration thickness from the real one.

CTDI₁₀₀ free‐in‐air measurements resulted in 13.03 mGy for 400 mAs at 120 kV, using the Ag filtration and a collimation of 4 cm. Using the Cu filtration with a collimation of 2 mm resulted in a measured CTDI₁₀₀ free‐in‐air of 246.2 mGy. The simulation for the Ag filtration resulted in 1.64·10⁻¹² mGy/photon, leading to 1.99·10¹⁰ photons/mAs. For the Cu filtration, the simulation resulted in 3.28·10⁻¹¹ mGy/photon, leading to 1.88·10¹⁰ photons/mAs.

3.2. Validation of the model

3.2.1. Bowtie filter

The relative doses at the lateral positions compared to the dose at the isocenter are shown in Figure 2. The simulated relative doses match the measured ones closely, with the maximum absolute difference being 2.3%.

FIGURE 2.

Relative dose as a function of lateral distance from the isocenter as measured and simulated. The maximum absolute difference of the relative dose is 2.3%.

3.2.2. Table transmission

The measured and simulated transmissions of the table were 74.5% and 75.2%, respectively.

3.2.3. CTDI measurements

Table 2 shows the results for the measured and simulated CTDI₁₀₀ for the various positions in the phantom. The maximum deviation is 8.1%.

TABLE 2.

Measured and simulated CTDI₁₀₀ values at the various positions in the 32 cm CTDI phantom.

	Ag filter—Tube rotating			Cu filter—Tube@top			Cu filter—Tube@lateral			Cu filter—Tube@bottom
Position	Meas. [mGy]	Sim. [mGy]	Dev. [%]	Meas. [mGy]	Sim. [mGy]	Dev. [%]	Meas. [mGy]	Sim. [mGy]	Dev. [%]	Meas. [mGy]	Sim. [mGy]	Dev. [%]
12	6.703	6.803	1.5	526.0	515.9	−1.9	29.41	28.85	−1.9	2.904	2.753	−5.2
3	6.571	6.687	1.8	29.12	29.09	−0.1	526.8	514.0	−2.4	24.50	25.16	2.7
6	6.016	6.139	2.0	3.933	3.613	−8.1	33.63	33.28	−1.1	482.37	465.8	−3.4
9	6.578	6.643	1.0	29.15	29.03	−0.4	3.577	3.465	−3.3	24.29	25.18	3.7
Center	3.812	3.789	−0.6	65.75	62.99	−4.2	66.05	63.36	−4.1	55.33	53.28	−3.7

3.3. Dose estimation in human models

The assignment of the various voxel designations to the tissues mentioned in table 3 of ICRP Report 103 can be found in the Supplementary information.

Figure 3 shows the relative energy distribution of the various localizer strategies. As expected, the dose distribution of the 3D localizer is more homogeneous compared to the other localizer strategies.

FIGURE 3.

Relative energy distribution in one slice of the ICRP adult male computational phantom for various localizer acquisitions strategies. Energy distribution for the rotational 3D localizer is more homogenous compared to the other localizer strategies.

Figure 4 shows the absorbed organ doses for selected organs of interest for the chest localizer acquisitions, while Figure 5 shows the effective dose, all as a function of BMI. Figures 6 and 7 show the same for the abdomen and pelvis localizer acquisitions, with different organs being included in Figure 6 compared to Figure 4. The results for the combined chest, abdomen, and pelvis localizers can be found in the Supplementary information. Since localizer acquisitions are fixed‐dose acquisitions, the trend is that the organ and effective dose are lower for higher BMIs. Also, the scans with PA acquisition have, in general, a lower organ dose and lower effective dose compared to scans with an AP acquisition. This is because most radiosensitive organs are positioned anteriorly in the human body and due to the attenuation of the table, resulting in a lower dose to these organs while the incident dose is the same for PA and AP acquisitions. The organ and effective dose for the 3D localizer acquisition is lower than for both the combined AP+LAT and PA+LAT localizers.

FIGURE 4.

Organ doses for the various phantoms using different localizer acquisition techniques for the chest region. Shown are the most radiosensitive organs in the chest region: red bone marrow (RBM), lung, breast, and stomach.

FIGURE 5.

Effective dose for the various phantoms using different localizer acquisition techniques in the chest region.

FIGURE 6.

Organ doses for the various phantoms using different localizer acquisition techniques for the abdomen and pelvis region. Shown are the most radiosensitive organs in the chest region: red bone marrow (RBM), colon, gonads, and stomach.

FIGURE 7.

Effective dose for the various phantoms using different localizer acquisition techniques in the abdomen and pelvis region.

For the chest acquisition, compared to our current clinical localizer strategy (AP+LAT), the PA+LAT would result in a 7 ± 2% effective dose reduction (mean ± standard deviation), while the 3D localizer would result in a 21 ± 3% effective dose reduction. Using AP or PA only would result in 69 ± 2% and 76 ± 2% effective dose reduction, respectively.

For the abdomen acquisition, PA+LAT would result in a 6 ± 2% effective dose reduction, while the 3D localizer would result in a 20 ± 5% reduction. Using AP or PA only would result in 69 ± 5% and 76 ± 4% effective dose reduction, respectively.

The data for all organs, for the combined chest and abdomen/pelvis acquisitions, as well as the doses per DLP instead of per full acquisition are available in the Supplementary information. The uncertainties for the organ doses for organs within the scan range as well as for the effective dose are all below 1%.

The results shown in these graphs shown separately for pediatric and adult patients can be found in the Supplementary information.

4. DISCUSSION

In this study, it was found that, as expected, the PA‐only localizer results in the lowest organ dose for the most radiosensitive organs and the lowest effective dose. This is because most radiosensitive organs are positioned more to the anterior of the human body and by the attenuation of the table in combination with using the same dose for PA and AP acquisitions. The 3D localizer results in a lower organ and effective dose compared to any combined two‐2D localizer acquisitions. Of course, organ and effective doses depend on the angle of the 2D localizer. Therefore, only comparing CTDI_vol,spr or DLP_spr for a given localizer acquisition is insufficient to determine if another localizer strategy would result in lower patient doses.

CT systems use different localizer strategies. Using only one localizer acquisition is, in terms of dose, the most efficient. However, this might make it harder to determine the correct exposure parameters if the patient is not correctly centered. Several studies have shown that underexposure or overexposure might occur in this situation. ¹² , ¹³ , ¹⁴ Two 2D localizers or a 3D localizer give more information about the local attenuation of the patient and may therefore result in better information for the automatic tube current modulation.

The newly introduced 3D localizer results, in general, in lower organ doses and effective doses compared to the currently‐used AP+LAT localizers. In addition, with these, the CT system can perform automatic landmark detection to automatically determine the scan range needed, which might be harder with 2D localizers, if at all possible. This automated range setting may result in more consistent imaging of patients.

Two previous studies estimated organ and effective doses for 2D localizers using Monte Carlo simulations. ¹ , ² These estimations were both done for Siemens CT systems having different filtration, bowtie filter, geometry, and exposure. Therefore, it is hard to compare the results quantitatively. For example, for chest localizers, Schmidt et al. reported an effective dose of 0.39, 0.25, and 0.23 mSv, for the AP, PA, and LAT localizer, respectively.1 These values are substantially higher than those reported in this study. Hoye et al. reported lower effective doses of 0.17, 0.08, and 0.09 mSv, for the AP, PA, and LAT localizer, respectively.2 These values are more in line with those of the current study for the AP and PA acquisition, but the LAT is about twice as high in the current study. Nevertheless, those two studies showed, as this one, that a PA localizer results in a lower effective dose than a localizer in the AP direction. For most organ doses and for the effective dose, the dose of the LAT localizer is in between the PA and AP localizer in all three studies, if the same exposure is used for the LAT localizer as for the AP/PA localizers.

In the EU, the median DLP for chest acquisitions with an indication for lung cancer diagnosis are as low as 130 mGy∙cm. ¹⁵ Using a generic conversion factor of 0.015 mSv/mGy∙cm, this results in an effective dose of about 2 mSv. ¹⁶ Depending on the localizer strategy, the localizer acquisition can take up to 20% of this dose. For the abdomen, the median DLP in the EU is 325 mGy∙cm, leading to an effective dose of about 5 mSv, resulting in localizer acquisitions that can take up to 10% of the total dose given to the patient. Therefore, it is important to include localizer acquisitions in patient dose estimates.

This study has several limitations. First, only a limited number of computational phantoms were included. However, since the results seem to be consistent over a relatively wide range of BMIs and the two types of computational phantoms (ICRP and XCAT), we are confident that the number of phantoms taken is sufficient to compare the different localizer strategies. Outside the BMI range used in this study (relative) results may differ. Another limitation is that only one specific CT system was simulated. This also means that for the combined PA/AP+LAT localizers, the exposure ratio of the PA/AP (30 mA) versus the LAT (100 mA) localizer of about 1/3, as mentioned above, was used. Other manufacturers seem to keep the CTDI or PA/AP versus LAT localizers the same. This would influence the results significantly. Of course, the interested reader can estimate the total combined dose for PA/AP + LAT acquisitions with any exposure for the two localizer acquisitions using the results included in the Supplementary information.

5. CONCLUSIONS

Using a PA localizer results in lower or equivalent organ dose in the most radiosensitive organs, and a lower effective dose, compared to an AP localizer for both chest and the combined abdomen and pelvis region. In addition, compared to a two‐2D localizer strategy, the 3D localizer results in a lower effective dose in both the chest and the combined abdomen and pelvis region.

CONFLICT OF INTEREST STATEMENT

Ioannis Sechopoulos has research agreements with Siemens Healthcare, Canon Medical, ScreenPoint Medical, Sectra Benelux, Volpara Healthcare, Lunit, iCAD, and a speaker agreement with Siemens Healthcare.

Supporting information

Supporting information

Supporting information

Oostveen LJ, Tunissen S, Sechopoulos I. Comparing organ and effective dose of various CT localizer acquisition strategies: A Monte Carlo study. Med Phys. 2025;52:576–584. 10.1002/mp.17447