© M. Sadeghi Moghadam and M. Keshtkar, Published by EDP Sciences, 2025

1 Introduction

In recent years, the use of computed tomography (CT) scans has become a crucial tool in diagnosing various medical conditions in pediatric populations. Specifically, abdomen-pelvis CT scans have proven to be invaluable in detecting and evaluating a wide range of diseases (Mwinyogle et al., 2020). However, as the utilization of pediatric CT scans has increased, concerns regarding the potential cancer risk associated with this imaging modality have emerged (Fahimi et al., 2012). The reason for this is that children’s developing tissues are much more susceptible to radiation-induced cancer than those of adults because their cells divide at a faster rate. Additionally, children have a longer lifetime ahead of them, which means they have more chances to develop radiation-related cancers. As a result, the likelihood of cancer occurring due to radiation exposure is greater in children (Miglioretti et al., 2013).

A variety of studies have been conducted to investigate the potential carcinogenic effects of ionizing radiation exposure from CT scans. One study by Pearce et al. (2012) analyzed a large cohort of patients and found a definitive association between radiation exposure from CT scans during childhood and an increased risk of developing cancer later in life. The authors estimated that the utilization of CT scans in children, with cumulative doses of approximately 50 mGy, could potentially increase the risk of developing leukemia by nearly threefold. Similarly, doses of around 60 mGy might triple the risk of brain cancer. These findings raised considerable concerns, particularly given the higher sensitivity of children to ionizing radiation.

Bosch (de Basea et al., 2018) conducted a research study with the objective of determining the approximate number of cancer cases caused by radiation exposure from CT scans in children and young adults in Spain during the year 2013. According to their calculations, a total of 168.6 cancer cases are expected to occur over a lifetime as a result of exposure to ionizing radiation from 105,802 CT scans. The risks of developing cancer per 100,000 patients exposed were found to be highest for breast and lung cancer.

To address the mounting concerns regarding cancer risk, various strategies and techniques have been proposed to reduce radiation exposure in CT scans. Low-dose CT techniques, such as iterative reconstruction algorithms (Lu et al., 2023), automatic exposure control (Wallace et al., 2015), and tube current modulation (Sookpeng et al., 2021), have been proven effective in reducing radiation dose while maintaining image quality. Additionally, radiologists are encouraged to follow the as-low-as-reasonably-achievable (ALARA) principle when prescribing and performing CT scans in children (Toma et al., 2019).

Despite efforts to minimize radiation exposure, the potential for cancer induction from pediatric CT scans remains a concern for both healthcare providers and parents. Consequently, the need for accurate assessment and communication of the potential risks associated with this imaging modality is essential (Kiani and Chaparian, 2023).

It is important to mention that a notable correlation exists between the amount of radiation received and the quality of diagnostic images. Additionally, when attempting to decrease the radiation dose for patients, adjusting the scan parameters can be helpful. However, decreasing these parameters may result in an elevation of image noise, which can have a detrimental impact on the quality of the image (Yel et al., 2019).

Henceforth, it is important to include details about the image quality when presenting radiation dose values and associated risks. The objective of this study is to assess the organ doses, effective dose, image quality, and the potential risk of developing radiation-related cancer in children undergoing CT scans of the abdomen-pelvis.

2 Materials and methods

2.1 Study design

Our institution’s ethics committee granted approval for the study, eliminating the need for consent forms. The research involved 219 pediatric patients who underwent an abdomen-pelvis CT examination between May 2021 and December 2023. These patients were categorized into four age groups: G1 (<1 year old), G2 (1–4 years old), G2 (5–9 years old), and G3 (10–14 years old). Data regarding the patients’ demographic information were collected using the picture archiving and communication system (PACS).

2.2 Collection of dose parameters

All individuals received imaging using a 16 slice CT scanner (Siemens, Germany). The indications for abdomen-pelvis CT were abdominal and pelvic pain, as well as trauma. All CT examinations were performed during the venous phase after contrast injection at a dose of 1.5 mL/kg. The dose report page was utilized to extract the volumetric CT dose index (CTDI_vol), dose length product (DLP), and the parameters for CT protocols, which were obtained by reviewing the PACS.

In this study also size specific dose estimate (SSDE) was calculated. SSDE is used to determine the radiation exposure considering the variation in patient sizes, as different patients may require different radiation doses due to their body size. The formula for SSDE is as follows (Medicine, 2011):

$$\text{SSDE}={\text{CTDI}}_{\text{vol}}\times \left(\text{patient}\hspace{0.17em}\text{size}\hspace{0.17em}\text{correction}\hspace{0.17em}\text{factor}\right)$$(1)

where CTDI_vol represents the amount of radiation emitted by the X-ray tube as it moves completely around the patient’s body. Patient size correction factor is a factor that accounts for the differences in patient size. The conversion factor is determined by the effective diameter and CTDI_vol stated in the AAPM 204 report on SSDE (Medicine, 2011). The measurement for the effective diameter of a CT scan of the abdomen-pelvis region was performed by determining the distance between the anterior-posterior (AP) and lateral (Lat) sides of the middle slice axial images. The AAPM 204 report were followed, which included using equation (2) for calculating the effective diameter based on 32 cm phantom and Table 1.

$$effectivediameter=\sqrt{AP\times Lat}$$(2)

2.3 Collection of organ doses and effective dose

The National Cancer Institute dosimetry system for CT (NCICT) software version 3.0 was utilized to estimate the doses of radiation received by organs, measured in millisieverts (mSv), by utilizing tissue weighting factors from the international commission on Radiological protection (ICRP) 103. Additionally, the software was also used to calculate the effective dose, also measured in mSv. The software offers the calculation of organ doses using the ICRP reference pediatric and adult phantoms as well as other measures of dose such as effective dose. These calculations are easily accessible through the graphical user interface, which also includes a feature for batch calculations. Overall, NCICT is a valuable tool for researchers seeking accurate estimations of organ doses in pediatric and adult patients undergoing CT examinations (Lee et al., 2015).

2.4 Cancer risk estimation

Estimation of cancer incidence risks were conducted according to the Biological Effects of Ionizing Radiation (BEIR) VII report (Council, 2006). This report is a comprehensive scientific study conducted by the National Research Council’s committee in the United States. The report evaluates the health risks associated with exposure to ionizing radiation. The BEIR VII report incorporates dose-response models, epidemiological studies, and experimental evidence to estimate cancer risks associated with radiation exposure. It also considers various factors that influence radiation sensitivity, such as age, gender (Council, 2006).

Cancer risks were calculated for leukemia, all solid cancers and all cancers without any differentiation.

2.5 Image quality assessment

The image quality was evaluated in this research by measuring the level of noise present, as well as the signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). To assess the image quality, the CT scans that were reconstructed with a thickness of 5 mm were utilized.

The placement of circular region of interests (ROIs) was done manually in the right hepatic lobe. The area of these ROIs was approximately 100 ± 10 mm². In addition, ROIs were also placed in the aorta, ensuring that they covered at least two-thirds of the vessel’s circumference while avoiding its walls. Similarly, ROIs were placed in the right paraspinal muscle, with an area of 100 ± 10 mm². To determine the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of the liver and aorta, the obtained values were used in the following equations (Mello-Amoedo et al., 2018):

$$SNR=\frac{CT{N}_o}{{\sigma }_0}$$(3)

$$CNR=\frac{\left(CT{N}_o-CT{N}_m\right)}{{\sigma }_m}$$(4)

where CTN_o, CTN_m, σ₀ and σ_m are mean CT number of organ of interest, mean CT number of paraspinal muscle, standard deviation (SD) of organ of interest, SD of paraspinal muscle, respectively.

2.6 Statistical analysis

The analysis of data was conducted using SPSS software (version 23). Mean and SD were utilized to represent all values. Additionally, the Mann–Whitney test was applied to compare the means of continuous variables between the two groups. One-way analysis of variance (ANOVA) was used to compare four groups in image quality assessment and dose parameters. The Spearman correlation analysis was used to assess the relationship between two variables. In order to consider statistical significance, any differences with a p-value below 0.05 were taken into account.

2.7 Use of AI in the writing process

The authors declare the use of assistive artificial intelligence, specifically OpenAI’s ChatGPT-4, in preparing this manuscript. This tool was used solely for text correction and language refinement in the writing process, primarily within the Introduction and Discussion sections. No AI assistance was involved in creating images, graphics, or tables.

3 Results

3.1 Patient’s demographic data and scanning parameters

Out of the 219 patients who took part in this study, 129 patients (58.90%) were males, and 90 patients (41.1%) were females. The mean age of males and females was 8.22 ± 3.56 and 8.75 ± 4.12, respectively. The values of effective diameter for <1, 1–4, 5–9, and 10–14 age groups were 13.87 ± 1.20, 14.22 ± 1.04, 16.84 ± 2.35, 18.66 ± 3.24 cm, respectively.

Table 1 shows CT scanning parameters for all age groups. With the exception of the youngest group, where the tube potential remained constant at 110 kVp, the lowest and highest tube potentials for the other age groups were 110 kVp and 130 kVp, respectively. The highest mean tube current was for the oldest age group. The tube rotation time and reconstructed slice thickness were constant for all age groups.

3.2 Radiation dose parameters and organ doses

The results of CTDI_vol, DLP, effective dose, and SSDE for four age groups are presented in Table 2. The results of one-way ANOVA showed that there were statistically significant differences between all age groups in terms of effective dose, DLP, and SSDE (p < 0.05). However, there was no statistically significant difference between the <1 and 1–4 groups in terms of CTDI_vol (p = 0.396).

The results of organ dose values from abdomen-pelvis CT examinations for each age group are summarized in Table 3. The colon received the highest organ dose of 8.72 mGy. After the colon, the small intestine, kidney, stomach wall, spleen, and pancreas also received the highest organ doses.

3.3 Cancer risk estimation

The values of cancer risks per 100,000 individuals induced by pediatric abdomen-pelvis CT scans for four age groups are shown in Table 4. As shown, for all investigated cancers, the youngest group had the highest cancer risks. The Pearson correlation test revealed that as the age of patients increased, there was a significant decline in all investigated cancer risks values (p < 0.05). The risk of all cancers was 75.38 per 100,000 for the <1 age group, and it decreased to 56.10 for the 10–14 age group.

Table 5 displays the cancer risks per 100,000 people by gender for four age groups. For all the investigated cancers, the cancer risks were statistically higher for females (p < 0.05).

3.4 Image quality

The results of image quality assessment in terms of noise, SNR, and CNR for four age groups are shown in table 6. There was no significant statistical difference in the mean noise of liver between various age groups (p > 0.05). However, the mean noise of the aorta showed a significant statistical difference between different age groups (p = 0.009). The findings revealed that the mean aorta noise in the youngest age group (<1 age group) was significantly lower compared to the 10–14 age group (p = 0.018), and the corresponding value in the 1–4 age group was lower than the 10–14 age group (p = 0.038).

Also, the mean noise of the paraspinal muscle showed a significant statistical difference between different age groups (p = 0.002). The findings revealed that the mean paraspinal muscle noise in the oldest age group (10–14 age group) was significantly higher compared to the <1 age group (P = 0.005), and the 1–4 age group (P = 0.010).

There was no significant statistical difference in the mean SNR of liver and aorta between various age groups (P > 0.05).

There was no statistically significant difference in the mean CNR of the aorta among different age groups (P = 0.136). The mean CNR of the liver showed a significant statistical difference between different age groups (P = 0.001). The results showed that the mean CNR of the liver in the oldest age group (10–14 age group) was significantly lower compared to the <1 age group (P = 0.004), and the 1–4 age group (P = 0.012).

A Spearman correlation analysis was performed to examine the relationship between radiation dose (CTDI_vol) and image quality (SNR in liver). Given that the data violated normality assumptions, as indicated by the Shapiro-Wilk test (P < 0.05), Spearman’s correlation was deemed appropriate. The analysis revealed no statistically significant correlation between radiation dose and image quality (Spearman’s r = –0.4, P = 0.75).

Table 7 provides the mean values of CTDI_vol, effective dose, and noise. These values are provided for the purpose of comparing our study with the study conducted by Muhammad et al. (2020). It should be noted that the noise values are for liver tissue.

4 Discussion

Pediatrics face an increased susceptibility to cancer development as a result of their exposure to ionizing radiation. Meanwhile, malignant tumors occurring in children have a greater fatality rate compared to cancers found in adults. Leukemia, thyroid, and breast cancers are frequently observed in children who have been exposed to ionizing radiation during their early years (Meulepas et al., 2019). Hence, it is highly important to accurately evaluate the radiation doses received by organs during pediatric CT scans and the associated risk of developing cancer.

The NCICT software was utilized to compute organ doses and effective doses for all patients involved in the research. In our study, the effective doses obtained for abdomen-pelvis CT scans across all age groups were found to be lower compared to several comparable studies (Gao et al., 2018; Miglioretti et al., 2013; Sulieman et al., 2022). In this study, the calculated mean effective dose for children in the age group <1 was 1.54 mSv, while Gao et al. (2018) reported a value of 2.8 mSv for the same age group. Also, the calculated mean effective dose in this study in the age group 10-14 was 4.05 mSv, while Miglioretti et al. (2013) and Sulieman et al. (2022) reported 14.8 and 7.2 mSv for the same age group, respectively.

Regarding the CTDI_vol, we observed no significant difference between the <1-year-old and 1-4-year-old groups, which may be attributed to the similar effective diameters in these age categories (13.87 ± 1.20 cm vs. 14.22 ± 1.04 cm) in our study. The use of the several different parameters for the same age group may have contributed to this lack of variation.

The colon was found to have the highest organ dose in this study, which is consistent with findings from Miglioretti et al. (2013), Tahmasebzadeh et al. (2021), and Gao et al. (2018) studies. The calculated organ doses in this study are lower than other studies (de Basea et al., 2018; Gao et al., 2018; Tahmasebzadeh et al., 2021).The mean calculated colon dose in this study for 10–14 age group was 8.72 mSv, while Miglioretti et al. (2013) and Tahmasebzadeh et al. (2021) reported 26.2 and 16.39 mSv, respectively. The variations in organ and effective doses found in our study compared to previous researches can be explained by disparities in the type of scanner used and the scan parameters employed.

In this study, the estimated risk of developing cancer from abdomen-pelvis CT scans was found to be comparatively lower across different age groups, when compared to other similar studies (Miglioretti et al., 2013; Tahmasebzadeh et al., 2021). In our study, the mean risk of developing all solid cancers among individuals aged 10–14 was approximately 5.3 per 10,000 people, whereas Miglioretti et al. (2013) reported a value of 40.3. Moreover, the mean risk of leukemia in the present study for the 10–14 age group was about 0.3 per 10,000 people, while Miglioretti et al. (2013) reported a value of 1.0. Also, the mean risk of leukemia in the present study for the 1–4 age group was about 0.41 per 10,000 people, while Tahmasebzadeh et al. (2021) reported a value of about 1.1. Also, the variations in estimated cancer risks found in our study compared to previous researches can be explained by disparities in the type of scanner used and the scan parameters employed. Also, the present study demonstrated the impact of patients’ age and gender on cancer risk results. In general, as patients grow older, the risk of cancer decreases because organs become less sensitive to radiation (Kiani and Chaparian, 2023).

Moreover, the risk of cancer was higher in young girls compared to young boys. This difference in cancer risk can be attributed to the fact that certain organs are more susceptible to radiation in females than males. In line with our study, according to the BEIR VII report, the risk of all cancers for a 5-year-old female exposed to a single dose of 0.1 Gy is approximately 1.86 times that of a 5-year-old male. The differences in cancer risk due to radiation exposure that we observed in our study, both in terms of age and gender, are in line with the findings of previous research (Kiani and Chaparian, 2023; Miglioretti et al., 2013; Su et al., 2014; Tahmasebzadeh et al., 2021).

Although the mean risk of leukemia in the present study due to abdomen-pelvis CT scans (4.16 per 100,000 females) is significantly lower than the baseline risk of leukemia in the absence of exposure (590 per 100,000 females) (Council, 2006), radiation dose reduction techniques, particularly in children, should not be disregarded. The cumulative effects of repeated exposures can still pose a long-term risk, and optimization of radiation protocols is crucial to minimize potential harm. This is especially important in pediatric populations, where sensitivity to radiation is higher, and even small reductions in dose could contribute to lowering overall cancer risks in the future.

In image quality assessment, the noise values of aorta and muscle, and CNR values of liver were different statistically significant across various age group. These differences may be attributed to variations in scan parameters. Image noise plays a crucial role in evaluating the image quality of CT examinations. Currently, there is a lack of sufficient references and guidelines specifically addressing the magnitude of noise in CT examinations. Consequently, the acceptable level of noise is influenced by the expertise of an experienced radiologist (Waite et al., 2019). Thanks to the progress of artificial intelligence, iterative reconstruction using deep learning-based images has made remarkable strides in reducing radiation dose and enhancing image quality. A recent study has presented evidence that deep learning algorithms can outperform commercially available iterative reconstruction from three leading vendors, resulting in superior image quality (Shan et al., 2019). No statistically significant correlation between radiation dose and image quality suggests that, within the range of dose levels used in this study, changes in radiation dose did not consistently correspond to improvements in image quality. These findings indicate that other factors, such as patient-specific characteristics (body mass index, effective diameter), may play a more critical role in determining image quality, independent of the radiation dose.

According to the study by Muhammad et al. (2020), the mean values of CTDI_vol and effective dose in all age groups were found to be higher compared to our study. However, the mean values of noise in our present study were observed to be lower when compared to the study conducted by Muhammad et al. (2020). Comparing image quality analysis with more other studies proved challenging in our research. The difficulty stemmed from various factors, which encompassed differences in scan parameters, CT scanner models and manufacturers, as well as variations in the reconstruction algorithms employed. These disparities made it problematic to establish direct comparisons with the image quality findings of other studies.

5 Conclusion

The findings of our study indicate that pediatric abdomen-pelvis CT examinations carry a potential risk of cancer. This risk becomes more significant in younger patients due to their heightened radiosensitivity. While the mean risk of leukemia from abdomen-pelvis CT scans in this study is lower than the baseline risk, radiation dose reduction remains essential, especially for pediatric patients. Given the cumulative effects of repeated exposures and children’s increased sensitivity to radiation, optimizing protocols to reduce dose is critical for minimizing potential long-term cancer risks. It is crucial for physicians to be mindful of these radiation risks and exercise caution when requesting a CT scan, ensuring that the advantages in terms of diagnosis surpass the risks associated with radiation exposure. Moreover, radiotechnologists can play a role in reducing radiation doses in the case of children, it involves optimizing scan parameters and using techniques to reduce radiation doses while still ensuring the required quality of diagnostic images.

Acknowledgments

The authors of this research would like to express their appreciation to the medical imaging Unit of Allameh Bohlool Gonabadi Hospital.

Funding

The Gonabad university of medical sciences provided funding for this project.

Conflicts of interest

The authors declare no conflict of interest in regards to this article.

Data availability statement

The data supporting this study's findings are available on request from the corresponding author, [M.keshtkar].

Author contribution statement

All stages of the research and manuscript writing was conducted equally by the Mohammad keshtkar, and Majid Sadeghi moghadam.

Ethics approval

The study received ethical approval from ethics committee of Gonabad university of medical sciences.

References

All Tables