A retrospective survey to establish institutional diagnostic reference levels for CT urography examinations based on clinical indications: preliminary results

Objective. To establish institutional diagnostic reference levels (IDRLs) based on clinical indications (CIs) for three- and four-phase computed tomography urography (CTU). Methods. Volumetric computed tomography dose index (CTDIvol), dose-length product (DLP), patients’ demographics, selected CIs like lithiasis, cancer, and other diseases, and protocols’ parameters were retrospectively recorded for 198 CTUs conducted on a Toshiba Aquilion Prime 80 scanner. Patients were categorised based on CIs and number of phases. These groups’ 75th percentiles of CTDIvol and DLP were proposed as IDRLs. The mean, median and IDRLs were compared with previously published values. Results. For the three-phase protocol, the CTDIvol (mGy) and DLP (mGy.cm) were 22.7/992 for the whole group, 23.4/992 for lithiasis, 22.8/1037 for cancer, and 21.2/981 for other diseases. The corresponding CTDIvol (mGy) and DLP (mGy.cm) values for the four-phase protocol were 28.6/1172, 30.6/1203, 27.3/1077, and 28.7/1252, respectively. A significant difference was found in CTDIvol and DLP between the two protocols, among the phases of three-phase (except cancer) and four-phase protocols (except DLP for other diseases), and in DLP between the second and third phases (except for cancer group). The results are comparable or lower than most studies published in the last decade. Conclusions. The CT technologist must be aware of the critical dose dependence on the scan length and the applied exposure parameters for each phase, according to the patient’s clinical background and the corresponding imaging anatomy, which must have been properly targeted by the competent radiologist. When clinically feasible, restricting the number of phases to three instead of four could remarkably reduce the patient’s radiation dose. CI-based IDRLs will serve as a baseline for comparison with CTU practice in other hospitals and could contribute to national DRL establishment. The awareness and knowledge of dose levels during CTU will prompt optimisation strategies in CT facilities.


Introduction
Computed tomography (CT) is employed in various diagnostic applications to detect tumours and guide interventional procedures [1]. CT urography (CTU) refers to a multi-phase CT scan, including postcontrast excretory phase imaging, to examine the urinary tract from the kidneys to the urethra [2,3]. Clinical indications (CIs) for CTU are the investigation of gross hematuria or persistent microhematuria, renal and urothelial masses, obstructive uropathy, strictures, congenital collecting system abnormalities and pre-operative planning in case of complex renal stones or nephron-sparing surgery [3][4][5][6][7].
Despite the relatively higher patient radiation dose, CTU has displaced the conventional modality for diagnosing urinary tract abnormalities, the intravenous urography (IVU) [8,9]. Nowadays, the two most commonly used CTU protocols are the singleand the split-bolus, including a non-contrast phase and one or more phases after iodinated contrast medium injection [3,[10][11][12]. The non-contrast phase detects the urinary tract calculi, providing baseline attenuation values for characterising the lesions. The optional corticomedullary and the nephrographic phases (20-30 s and 80-120 s after contrast medium administration) detect tumours in the kidneys, collecting system, ureters, and urinary bladder, providing information on vascular anatomic and renal function. During the excretory phase (5-15 min post-injection), renal excretion of contrast medium helps detect any abnormalities in the collecting system, ureters, and urinary bladder, assisting the localisation of stenoses, malignancies, etc [13][14][15].
However, the CTU examination is not represented by a standardised protocol as observed in the literature [2,[10][11][12][13][14][15][16][17][18]. The technique variation amongst different imaging departments is mainly due to the variations in the CTU protocols, which depend on patient cohorts, CIs, and competent physician preferences. The CT protocols are generally individualised on a patient by the CT equipment using automatic tube current modulation (ATCM), tube voltage selection, etc or by the CT technologist by modifying the standard acquisition protocol. Some institutions initially used the splitbolus dual-phase technique in all patients and later performed triple-phase scanning following suboptimal excretory phase imaging [16]. Other studies proposed using a single unenhanced CT scan in riskfree patients (younger patients with microhematuria) [17]. Some institutions even use a triple-bolus technique providing images during unenhanced, corticomedullary, nephrographic and excretory phases [3]. This technique simultaneously demonstrates the enhancement of the arteries, renal parenchyma, and the collecting system [18].
Studies by Nawfel et al [19] and Silverman et al [20] revealed that a standard three-phase CTU increases radiation dose approximately 1.5 times compared to a conventional IVU. Dose reduction is mainly obtained by limiting the scan duration, number of phases, and application of lower exposure parameters. Split-bolus techniques are reserved for younger patients to minimise radiation dose [3], and the iterative reconstruction (IR) feature provides additional ways to reduce patients' radiation dose [21]. Low-dose CT gradually becomes a trustworthy alternative method in cases where high contrast can be naturally achieved, such as renal stone evaluation [22].
Besides the mandatory justification of a CTU examination, continuous efforts should be made to optimise radiation protection during these procedures. The common goal of established regulations is to keep the patient dose as low as reasonably achievable (ALARA) [23,24] based on the concept of Diagnostic Reference Levels (DRLs) [25]. Our team has recently worked on establishing Institutional DRLs (IDRLs) in standard CT and CT angiography examinations [26,27] as an initial step towards planning optimisation strategies concerning these examinations. However, there are no studies in Greece reporting DRLs regarding CTUs. Additionally, most existing DRLs have been established based on anatomical locations. However, this has some limitations. For example, one could have several CIs with different protocols corresponding to different exposure levels [28], image quality parameters and scan lengths, which could exhibit different DRLs.
The purpose of this retrospective study was to report volume computed tomography index (CTDI vol ) and dose-length product (DLP) IDRLs of CTU examinations performed with two different techniques (a three-phase split-bolus protocol and a four-phase single-bolus protocol). IDRLs were established per the number of phases and selected CIs. In addition, the proposed IDRLs were compared to those published by several international surveys.

Materials and methods
Patients' sample and CTU examinations Patients demographics (gender and age) and weight were recorded for 198 selected CTU examinations performed at the University Hospital of Patras' CT Department from January 2016 to November 2021. Ethics approval was obtained by the Institution's ethics committee. However, because of the study's retrospective nature, patients' consent was not required by the Hospitals' Institutional Review Board.
Patient's demographics (gender, age) and weight, per phase, per selected CI for the three-phase and four-phase protocols, and as a total, are presented in table 1.
We retrospectively reviewed certain cases of CTU matching our Department's criteria of three-and four-phase protocols, requested from the urology clinic of our Institution, following the corresponding international guidelines [29,30]. An effort was made to collect a sample including approximately the same number of patients regarding the sex and applied protocol. The main exclusion criteria were: no paediatric patients, no examinations with diagnostic uncertainty, and no examinations in which the initial protocol was altered (an additional phase was intermediately added to the three-phase protocol) due to improper image quality for diagnosis. The final sample consisted of 119 CTUs performed in patients with complex urinary stones requiring imaging planning prior to percutaneous or endoscopic treatment (henceforth briefly referred to as lithiasis); 26 patients with cancer, including renal, urinary tract, and bladder cancer; and 53 patients with other diseases of the urinary system, such as hematuria, hydronephrosis, and obstructive uropathy.
The radiologists in our Department follow the ALARA principle and always consider the patient's age and/or clinical conditions/indications before they select the three-or four-phase protocol to perform the CTU examination. The examinations included in the present study consist of either a split-bolus threephase protocol (95 patients) or a single-bolus fourphase protocol (103 patients). The CTU protocol is selected according to the CI (cancer, lithiasis, or other diseases), depending on the patient's clinical background and the radiologists' demands. For cancer detection and characterisation, a single-bolus fourphase protocol was mainly selected. In contrast, the split-bolus three-phase protocol was preferred for lithiasis and other diseases. Also, the three-phase protocol is recommended for younger patients (age 40 years old) to ensure low dose levels [3], considering their longer life expectancy and the ALARA principle. However, this protocol is less sensitive to detecting smaller renal cell carcinomas than the four-phase one. The threshold of 40 years is selected as an approximate threshold when the radiologists in our Department prescribe the three-phase protocol. The three-phase (split-bolus) protocol is applied in patients less than 40 years old since, according to Cheng et al [3], this population has a low incidence of pathologies not detectable on unenhanced CT alone, likely due to the low pretest probability for urothelial malignancy. For those patients whose excretory function is unsatisfactory or whose images proved inappropriate for accurate diagnosis a delayed fourth phase is added during the ongoing procedure. These cases were excluded from the sample to match the three-and four-phase protocols with split-and single-bolus techniques, respectively. On the other hand, the four-phase protocol produces a more accurate diagnosis at the cost of a higher radiation dose [3]. Therefore, it is usually applied to older patients due to the relatively lower radiation risk. Nevertheless, selecting the protocol to be applied is always performed individually to maximise the patient benefit/risk ratio.
Regarding the split-bolus three-phase protocol, a non-contrast abdomen scan extending from the diaphragm level to that of the pubic symphysis is performed. Non-contrast images are essential for detecting renal fat-rich, cystic or calcified lesions and as a reference for comparison with the structures during the following phases. Then, an intravenous injection of contrast medium (75 ml at an infusion rate of 3.5 ml s −1 ) is performed, followed by a second injection (55 ml with the same infusion rate), at a 50 s time interval. Scanning is performed at 90 s from the beginning of the first injection. Thus, a combined arterial and nephrographic phase is achieved. It is followed by the excretory phase, and the last scanning is performed about 8 min post-injection. In case of dilated urinary tract, excretory phase was acquired about 15-30 min post injection.
As far as the single-bolus four-phase protocol is concerned, the initial unenhanced abdomen scan is similar to the split-bolus protocol. Then, it is followed by a complete '3-phase kidneys' scan: an ancillary arterial (corticomedullary) phase that begins 10 s after the amount of contrast medium in the abdominal aorta has reached a threshold enhancement of 180 Hounsfield Units (bolus triggering technique); a nephrographic and an excretory phase, at 90 s and 8 min, after contrast medium injection, respectively. These time intervals may vary among patients depending on their cardiac function. The scan length extends from the kidneys to the bladder during the contrastenhanced phases, except for the arterial phase, which is restricted to the kidneys region to avoid excessive radiation. The clinical parameters and use of each protocol are summarised in tables 2 and 3.
The tube voltage applied in the initial pre-contrast scan both for the three-and four-phase protocols was 100 kVp for underweight and normal-weight patients and 120 kVp for overweight and obese patients. For all contrast-enhanced phases of both protocols, the tube voltage was 120 kVp, independent of the patient size. The tube current for all phases of both protocols was altered through ATCM. The technical parameters of the two acquisition protocols are presented in table 4.

CT system, dosimetric indices, and quality control
The examinations were conducted on a Toshiba Aquilion Prime 80 (Canon Medical Systems Corporation, Japan) CT scanner. For each patient, exposure parameters (tube voltage and tube current) and dosimetric indices (CTDI vol and DLP) were retrospectively recorded from the picture archiving and communications system of the Hospital. All the scans were reconstructed using the adaptive iterative dose reduction (AIDR) algorithm. In addition, the technical parameters included the scan type, rotation time, nominal collimation width, nominal slice thickness, reconstruction slice thickness, kernel and pitch factor (table 4).
For each CI, patient radiation dose was assessed utilising the recorded CTDI vol and DLP indices. CTDI vol describes the radiation output of a CT scanner [31]. It depends on the CT scanner type and the exposure parameters (tube voltage and tube current). For examinations performed in the torso, the estimation of CTDI vol is based on a cylindrical phantom with a 32-cm diameter (body phantom) [32]. The DLP refers to the patient's overall incident x-ray energy and is equal to the CTDI vol multiplied by the total scan length. These indices do not account for the patient's actual size and body composition; thus, they represent approximately the dose absorbed by the patient [33]. Clinical-based IDRLs were established in the 75th percentiles of the CTDI vol per phase and DLP per examination [25]. In this study, we defined IDRLs for CTU examinations at 75th percentile of the distribution of the medians, for one x-ray facility, taking into consideration that this is the largest public Hospital that covers a large geographical region in Western Greece, and a large number of specialised CTU examinations are performed for which no national DRL values exist to serve as an aid for optimisation [25]. IDRLs are also established to account for the dose reduction that could be achieved through the application of dose reduction techniques (lower kVp during ph-1, ATCM, and AIRD 3D). Thus, these values could contribute to further optimisation in our Hospital by providing a local comparator for the Department's quality assurance program in the future [25]. Additionally, IDRL values were accompanied by information regarding the patient data collected, the details of the specific examination, and technical parameters such as detector technology, detector configuration, and the image reconstruction algorithm, as recommended by ICRP 135 [35]. Comparison with corresponding values of the international literature has been made using the median and the 75th percentiles of the total CTDI vol and total DLP. The respective mean values have also been reported to facilitate comparison with the other studies.
Quality control tests were performed regularly throughout the study period by the Institution's Medical Physics Department to confirm the scanner's optimal performance in terms of image quality and radiation output [34].

Statistical analysis
The mean, median, interquartile range and 95% confidence interval of the mean CTDI vol and DLP values obtained with the three-and four-phase protocols and the three CIs were estimated. The mean, median and IDRLs for CTDI vol and DLP values were compared with corresponding values from the   The CTDI vol represents the scanner's radiation output for specific irradiated anatomy (kidneysureters-urinary bladder), CT protocol (threeor fourphase), and each phase. Therefore, it is not expected to vary significantly among CIs, as confirmed by the results presented in figures 1 and 2, since homogeneous patients' samples of body weight are involved (table 1). For example, the mean CTDI vol values for the first phase (ph-1) were 5.9/5.5/5.7 mGy (three-phase protocol) and 6.1/5.8/6.5 mGy (four-phase protocol) for lithiasis, cancer, and other diseases, respectively. However, significant differences among the phases (ph-1/ph-2/ph-3) were exhibited in the three-phase protocol for lithiasis, other diseases and as a total, or among phases (ph-1/ph-2/ph-3/ph-4) of the fourphase protocol for all CIs and as a total, attributed to the different exposure parameters. Specifically, both protocols used a tube voltage of 100 kVp during ph-1 for underweight and normal-weight patients, while 120 kVp was selected for overweight and obese patients and all the other phases. The tube voltage was selected according to the radiologists' demands, correlated to the patient's age and clinical background. Additionally, the tube current was altered through the ATCM to maintain consistent image quality according to the anatomical characteristics (table 1). The influence of ATCM is apparent when comparing male to female patients since, for all phases of each protocol, the tube current values were higher for males than females for the same phase. Specifically, the mean current applied during ph-1, ph-2, and ph-3 of the 3-phase protocol was 154, 142, and 136 mA for males and 115, 108, and 104 mA for females, respectively. Similarly, the mean tube current values during ph-1, ph-2, ph-3, and ph-4 of the 4-phase protocol was 163, 141, 128, and 138 mA for males and 134, 111, 108, and 118 mA for females, respectively. No significant differences were found when comparing CTDI vol values among the phases (ph-1/ph-2/ ph-3) of the two protocols for all CIs. However, the additional phase of the four-phase protocol resulted in a significant increase in the total CTDI vol compared to the three-phase protocol (mean values: 26.3 versus 18.7 mGy).
Regarding the CIs, no significant difference was found for CTDI vol and DLP values for each phase and as a total (Kruskal-Wallis test, p > 0.05) for both protocols (figures 1 and 2). Additionally, the total CTDI vol and DLP values differed significantly between threeand four-phase protocols for the total group of patients (Mann-Whitney test, p < 0.0001). A significant difference in CIs was found only for DLP between ph-2 and ph-3 of lithiasis (Mann-Whitney test, p < 0.0001; p = 0.020); other diseases (Mann-Whitney test, p = 0.003; p = 0.012); as well as the total group of patients (Mann-Whitney test, p < 0.0001; p = 0.0004).
The scan lengths per phase, per CI and as a total for three-phase and four-phase protocols are presented in figure 3. A significant difference was found in scan length values among the phases of the three-phase protocol for lithiasis (Friedman test, p < 0.00001), other diseases (Friedman test, p = 0.00003) and the total group (Friedman test, p < 0.00001), but not for cancer patients (Friedman test, p = 0.647). For the four-phase protocol, a significant difference was found among the phases of all the groups studied (Friedman test, p < 0.00001). Regarding the CIs, no significant difference was found for scan length values for each phase (Kruskal-Wallis test, p > 0.05) for both protocols. Additionally, the total scan length values differed significantly between three-and four-phase protocols for the total group of patients (Mann-Whitney test, p < 0.0001). At the same time, regarding CIs, a significant difference was found for ph-2 and ph-3 of lithiasis (Mann-Whitney test, p < 0.0001), other diseases group (Mann-Whitney test, p = 0.001) and the total group of patients (Mann-Whitney test, p < 0.0001), and ph-3 of cancer patients (Mann-Whitney test, p = 0.029). At this point it should be noted that when quoting the CTDI vol of a multiphase examination, only the CTDI vol values for each phase should be presented (figures 1 and 2). However, in previous surveys (table 5) a total mean/median CTDI vol calculated as the sum of all phases is reported to provide a comparison among multiphase examinations. Additionally, it should be noted that without clear information on what each phase is covered (in terms of the body part and the use of contrast medium), the mean/median CTDI vol of all phases does not provide useful information. Thus, the IDRLs were established in terms of CTDI vol per phase and DLP per examination. Nevertheless, in table 5, the median and 75th percentiles of the total CTDI vol for all phases are presented to perform an approximate comparison with the results reported in previously published studies given that no information are reported in most of these studies regarding the scanned region and use of contrast medium.

Discussion
Although CTU is not typically indicated for the identification of urinary tract calculi according to the American Urological Association (AUA) guidelines [29], as a single unenhanced phase could be sufficient, stones are frequently detected on CTU when evaluating asymptomatic hematuria [3]. Indeed, many cases in the selected sample were imaged during urological preoperative treatment planning and diagnosed with complex stones, categorised under the general term 'lithiasis' for the purposes of the present study. Regarding the optimal initial imaging modality (CT, MRI, or ultrasound) to use for the evaluation of patients with suspected obstructive nephrolithiasis, guidelines differ among the American College of Radiology, AUA, and European Association of Urology [35].
This survey is the first conducted in our Hospital regarding establishing IDRLs in CTU examinations, based on the selected CIs and the number of phases. In Greece, the Greek Atomic Energy Commission (GAEC) has established DRLs for standard CT examinations; however, for CTU examinations, there are currently no national DRLs. Thus, the proposed IDRLs will contribute to establishing national DRLs and may be used as a baseline for optimising patient radiation protection in our Hospital and other Hospitals.
The CTDI vol values are significantly different among the phases of the two protocols for all CIs. However, the largest differences were found for other diseases (up to 14.8%), probably due to the different gender distribution between the two protocols (7/19 male/female ratio for the three-phase protocol and 17/10 for the four-phase protocol, respectively, table 1) and the different tube current values as defined by the ATCM (table 1).
The DLP is a dose metric considering the scanner's radiation output and the scan length. Although DLP includes CTDI dependence, different trends in DLP differences can emerge since the scan length may vary considerably (figure 3) regarding the differences among the patients' body habitus and clinical background, radiographer's and/or radiologist's experience, scanning preferences, and radiation protection training. The study's results indicate that DLP values did not vary significantly per CI for the three-or the . Histograms (mean, 95% confidence interval of the mean) and box plots (middle line: median, central box: interquartile range (25% to 75%), whisker: extends 1.5 times the interquartile range from the first and third quartile values, excluding any 'outside' (upper or lower quartile ±1.5 times the interquartile range) or 'far out' values (upper or lower quartile ±3 times the interquartile range), displayed as separate points) of the scan length values per CI, and each phase of the (a) three-phase and (b) four-phase protocols. Lithiasis refers to complex urinary stones requiring an imaging planning prior to percutaneous or endoscopic treatment, while other diseases include hematuria, hydronephrosis, and obstructive uropathy.   four-phase protocol (figures 1 and 2). However, significant variations were found between the phases of each protocol, except for cancer patients in the threephase protocol and patients with other diseases in the four-phase protocol, due to the differentiations in the exposure parameters (tube voltage and tube current,  table 3) and the scan lengths ( figure 3). Specifically, significant differences were found for ph-2 and ph-3 in the case of lithiasis, other diseases and as a total (figures 1 and 2). The DLP values were lower in ph-2 and higher in ph-3 of the four-phase compared to the corresponding phases of the three-phase protocol due to the different scan lengths for these phases (figure 3) between the two protocols (ph-2 of the four-phase protocol is extended only to the kidneys and does not cover the whole abdominopelvic region, as in ph-2 of the three-phase protocol, table 2). Also, the additional phase significantly increased the total DLP for the four-phase compared to the three-phase protocol (mean values: 1022 versus 780 mGy.cm, i.e., a percentage increase of 31%). Table 5 presents the proposed IDRLs based on the CI for the three-and four-phase CTU protocols, estimated on the 75th percentile of the total CTDI vol and DLP distributions. The average patient weight in the sample was 74.9 kg (table 1), per the recommendations for DRL establishment [25]. Even though all the other studies presented in table 5 reported mean and median values [13][14][15][36][37][38][39][40][41][42], a direct comparison with the proposed IDRLs may not be appropriate, mainly due to different CT scanner models and the absence of standardised CTU protocols according to the CIs. Nevertheless, the proposed IDRLs (table 5) and the median values (figures 1 and 2) for the threeand four-phase protocols are, in most cases, lower or comparable with the median values reported in the published studies. For the two-phase protocols, the total mean/median CTDI vol and the total mean/median DLP values range from 8 to 88 mGy and 571 to 1166 mGy.cm, respectively. For the three-phase protocols, the total mean/median CTDI vol and total mean/median DLP values range from 13.6 to 44.5 mGy and from 581 to 2320 mGy.cm, respectively; while for the four-phase protocols, the corresponding values range from 11 to 124 mGy and from 1957 to 2065 mGy.cm. To reduce this variation, it is important to homogenise the CTU protocols in terms of number of phases, exposure and technical parameters while maintaining acceptable image quality according to the clinical conditions. Wang et al [36] reported that the median CTDI vol and the DLP values were 44.5 mGy and 2320 mGy.cm for the three-phase protocol. These values are 175% and 254% higher than the median values in our study (figure 1). The above could be attributed to the higher tube voltage of 120 kVp for all patients and all phases instead of 100 kVp for underweight or normal-weight patients during the unenhanced phase in our study. In a recent multicentre study, important variations in CTU protocols regarding the tube voltage, tube current (fixed or ATCM), and filtered backprojection (FBP) or IR algorithms were reported [37]. The median CTDI vol and DLP values (14 mGy and 1793 mGy. cm, respectively) are 14% lower and 174% higher than the median values found in our study (figure 1). The three-phase protocol reported by Gifford et al [39] consisted of a tube voltage of 120 kVp in all phases, resulting in a mean DLP of 1324 mGy.cm, 70% higher than in our study (figure 1). Van der Molen et al [40] reported DLP on two groups based on the reconstruction algorithms. The use of an FBP algorithm and a tube voltage of 120 kVp resulted in a median DLP of 1382 mGy.cm, while the use of the AIDR 3D algorithm and a tube voltage of 120 kVp resulted in a median DLP of 581 mGy.cm, which is 11% lower than in our study (figure 1). In another study, Juri et al [41] reported mean CTDI vol values of 27.5 mGy for a normal-dose protocol and 10.7 mGy for a low-dose protocol utilising the AIDR 3D algorithm, which is 43% lower than in our study (figure 1). In both protocols, the tube current was adjusted (from 10 to 550 mA) through ATCM. Lee et al [42] reported a reduction in CTDI vol and DLP values up to 30% and 32%, for a three-phase protocol when the tube voltage is lowered from 120 to 100 kVp. Dahlman et al [15] used a lowdose three-phase protocol with a tube voltage of 120 kVp and effective tube current reduction during each phase, and reported a mean CTDI vol of 13.6 mGy, which is 27% lower than in our study ( figure 1). However, using a normal-dose protocol with a constant tube voltage of 120 kVp and effective tube load further reduced during successive phases resulted in a mean CTDI vol value of 23.3 mGy, 24% higher than this study (figure 1).
Regarding the four-phase protocols, only Gershan et al [37] and Abedi et al [14] reported dose values (table 5). Gershan et al [37] reported a median CTDI vol of 11 mGy and a median DLP of 2065 mGy.cm, 54% lower and 128% higher than in our study (figure 2). On the other hand, Abedi et al [14] reported mean CTDI vol and DLP values of 124 mGy and 1957 mGy. cm, which are 371% and 91% higher than in our study (figure 2). The above differences are attributed to the different number of phases and probably to the different scanning lengths, exposure parameters (tube voltage and tube current), or the reconstruction techniques utilised. However, no information regarding the above parameters is available [14,37].
Several authors have also reported the third quartile [38], median [36,37] and mean [13,14,39] values of a split-bolus two-phase protocol. For example, Gershan et al [37] reported low median CTDI vol (8 mGy) and DLP (740 mGy.cm) values incorporating data from 55 patients. When comparing the mean dose values of our three-and four-phase protocols (figures 1 and 2) with the two-phase published protocols, up to 370% and 234% difference was found for the CTDI vol , and up to 50% and 14% for the DLP, respectively. Nevertheless, motivated by a radiation protection culture, the radiologists in our Hospital applied a restricted scan length (figure 3) during ph-2 of the fourphase protocol to limit the extra dose due to the additional phase/s incorporated to improve diagnosis, compared to the most common selected protocols of two-or three-phases, for the split-and single-bolus technique, respectively. Additional studies associate dosimetric indices with the patient's weight without reporting total values [43] or providing dose values that refer exclusively to the excretory phase [44]. Other studies provided data regarding single-, five-, and sixphase protocols [37]. Furthermore, in another study with a dual-energy multi-detector CT scanner, a dose reduction of up to 55% was reported when omitting the actual unenhanced scans with the aid of virtual unenhanced reconstruction techniques [45].
Due to the relatively high radiation dose, it is recommended that radiologists, radiographers, and medical physicists make continuous efforts to ensure optimisation of the ratio between the patient benefit (successful diagnosis) and the potential health detriment due to excessive or needless radiation dose. Consequently, the principles of justification and ALARA [23] should be applied in CTU practice. Furthermore, within the framework of optimisation of radiation protection, there is the need for the application of standardised protocols according to CIs, depending on the individual patient referral and the imaging expectations, and highlighting the critical role of CT technologists' continuous education, while radiologists should focus on clinical rather than on anatomical DRLs.
Despite the adequate number of patients in the survey, the availability of dosimetric data, exposure conditions and clinical practice of other CT scanners and hospitals could have strengthened the results. Additionally, the lack of patient height values did not allow for a body mass index-based or a size-specific approach in the dose estimations, as recommended by the American Association of Physicists in Medicine [46]. Furthermore, image quality assessment was not considered, even though all the acquired images were of high diagnostic quality according to the radiologists. Regardless of these limitations, until the GAEC establishes national DRLs for CTU examinations, the IDRLs obtained in this survey could be a useful guide for the optimisation of radiation protection both in our Hospital and other Institutions. As future work, evaluating the effective dose and organ dose values could be an additional step towards optimisation of patient protection during CTU examinations.

Conclusions
This study proposes IDRLs for CTU examinations considering the number of phases and selected CIs. No statistically significant differences are found in IDRLs per CI for the three-and the four-phase protocols. However, statistically significant differences are observed among the separate phases, and between females and males, according to the scan lengths and the individual anatomic characteristics. Aiming to reduce the radiation burden originating from a CTU examination, our results confirm the critical role of the CT technologist regarding the appropriate scan length of each phase, according to the individual patient's clinical background and the corresponding imaging target. Additionally, restricting the number of phases when it is clinically feasible (applying a threeinstead of a four-phase protocol) significantly reduces the patient dose by 31%. Dose reduction techniques considering the individual anatomic characteristics, such as tube current modulation in all phases and lower tube voltage in the unenhanced phase, also favor the ALARA principle. Our study's mean, median and IDRL values are comparable to or lower than most previous studies in the last decade. A wide variation is observed in the CTDI vol and DLP values found in the literature, mainly depending on the scanners' type, number of phases and specific technical parameters of the acquisition protocols. The awareness and knowledge of radiation dose levels and homogenisation of the exposure and protocols' technical parameters according to the CIs during CTU examinations will prompt dose optimisation strategies in CT facilities.

Data availability statement
The data cannot be made publicly available upon publication because they are not available in a format that is sufficiently accessible or reusable by other researchers. The data that support the findings of this study are available upon reasonable request from the authors.