Clinical Low-Dose Computed Tomography Application for Urinary Calculi Detection Based on In Vitro Model

Article information

J. Radiat. Prot. Res. 2025;50(3):157-164

Publication date (electronic) : 2025 July 16

doi : https://doi.org/10.14407/jrpr.2024.00318

Yingdi Xie, Yangchun Gu, Jingshan Tao, Hailing Liu, Ke Zhu

Department of Radiology, Binhai County People’s Hospital, Yancheng, China

Corresponding author: Ke Zhu, Department of Radiology, Binhai County People’s Hospital, No. 299 Haibin Avenue, Binhai County, Yancheng 224500, China, E-mail: zhuke4863@163.com, https://orcid.org/0009-0001-6547-1765

Received 2024 October 24; Revised 2025 January 7; Accepted 2025 January 31.

Abstract

Introduction

Urinary calculi represent one of the most prevalent disorders associated with the urinary system, with a worldwide incidence ranging from 1% to 20% [1]. In the United States, the lifetime prevalence of urinary calculi is estimated at 10%–15% [2], while approximately 6.4% of patients in China are affected by this condition [3]. Due to rising rates of unhealthy diet and lifestyle choices, the incidence and recurrence of urinary calculi are increasing, with a recurrence rate of up to 52% within 10 years [4], particularly among children and young adults [5, 6]. Several methods are available for detecting urinary calculi. Historically, the primary imaging techniques included plain film of kidneys, ureters, and bladder (KUB), intravenous pyelography, and ultrasound (US). However, computed tomography (CT) has gained significant popularity for urinary calculi detection due to its rapid scanning capabilities, high resolution, and three-dimensional reconstruction [7]. Consequently, CT is now widely used for the imaging the diagnosis of urinary calculi. Despite its effectiveness, patients often undergo multiple CT scans for accurate diagnosis and disease evaluation. While conventional-dose computed tomography (CDCT) provides satisfactory diagnostic results, it poses considerable radiation risks [8]. Repeated CT exposures can potentially harm the human body and increase cancer risk [9]. To balance diagnostic quality with minimal radiation exposure, the medical research community must revisit this issue in line with the ‘as low as reasonably achievable (ALARA)’ principle [10].

Low-dose computed tomography (LDCT) has recently emerged as a method to reduce patient radiation exposure, in line with the ALARA principle. Initially, LDCT was primarily used for lung cancer screening [11, 12]. However, with its growing application, an increasing number of studies have investigated its efficacy in urology. Several approaches, such as decreasing tube current or using a tin filter, have been explored to reduce the CT radiation dose [13, 14]. Numerous studies have demonstrated that LDCT is an effective tool for diagnosing urinary calculi while reducing the effective radiation dose [14, 15]. However, radiation dose reduction is often accompanied by increased image noise and reduced image quality, which may negatively impact stone evaluation.

This study aimed to assess the feasibility of applying optimal low doses, derived from in vitro models, in clinical settings. The integration of a laboratory model with clinical validation strengthened the evidence for the feasibility and practicality of the optimal LDCT criteria in detecting urinary calculi in clinical practice.

Materials and Methods

1. In Vitro Model Scanning

1) In vitro modeling

Six distinct compositions of human urinary calculi—whedellite, uricacid, cystine, struvite, apatite, and brushite—were obtained from the urology department of the participating hospital between July and December 2022. Four sizes (1, 2, 4, and 7 mm) of each of the six urinary calculi were collected, resulting in a total of 24 stones (Fig. 1A). Meanwhile, a pork kidney with a smooth surface was acquired, submerged in normal saline (NaCl 0.9%), and manually compressed to eliminate internal gas as much as possible. The kidney was sliced along the renal pelvis and scanned to ensure the absence of scattered calcifications. Human urinary calculi of the same composition but varying sizes were then independently implanted into the renal pelvis (Fig. 1B and 1C). The experiments were conducted collaboratively by two radiology technologists, a specialist radiologist, and two urology residents. To ensure consistency, the procedure was repeated in a standardized manner for each stone.

Fig. 1

Pork kidney model. (A) Various compositions of human calculi, each available in four sizes: 1, 2, 4, and 7 mm, (B) 1 mm stone located in the renal pelvis, 2 mm stone in the renal pelvis, 4 mm stone in the lower calyx, and 7 mm stone in the upper calyx, (C) pork kidney during computed tomography scanning.

2) Scanning methods

CT scans were performed using a 64-slice multi-detector CT scanner (Optima CT660; GE Healthcare) with a pitch of 1.0, rotation time of 0.5 second, slice thickness of 5 mm, tube voltage of 120 kVp, and tube loading set at 120, 100, 80, 60, and 40 mAs.

3) Image analysis

All CT images were independently analyzed by two radiologists with 8 years of experience, who were blinded to the stone composition and scanning conditions. The primary endpoint was to determine whether the four stones of each composition were successfully detected under varying scanning conditions.

Results

2. Stone Detection in the In Vitro Model Scan

A total of 24 urinary calculi, with diameters of 1, 2, 4, and 7 mm, were detected using tube voltage of 120 kVp and tube loading of 120, 100, 80, 60, and 40 mAs, respectively. The detection rate was 100%. However, at 40 mAs, the image signal-to-noise ratio was poor, leading to use of 60 mAs in the clinical trial. Fig. 2 shows images of 1 mm stones of different composition scanned at 60 mAs.

Fig. 2

Computed tomography scan of stones performed at 60 mAs: (A) whedellite stone; (B) uricacid stone; (C) cystine stone; (D) struvite stone; (E) apatite stone; and (F) brushite stone. The white arrows show the 1 mm stone.

3. Stone Diagnoses in Clinical Studies

As shown in Table 1, no statistically significant differences were observed between CDCT and LDCT in terms of stone location and number. Both CDCT and LDCT detected a total of 285 stones, with 148 stones located in the kidney and 137 in the ureter. The maximum diameter of the stones ranged from 1 mm to 20 mm. The ICC analysis demonstrated a strong correlation between the length, width, and height measurements of stones obtained by CDCT and LDCT (Table 2), indicating that both methods exhibited excellent consistency in assessing the size of urinary calculi. Fig. 3 presents the reconstructed images from both CDCT and LDCT scans.

Table 1

Comparing the Stones Detection Rate between CDCT and LDCT

Table 2

ICC Analysis of Stone Size Measured by CDCT and LDCT

Fig. 3

Images of urinary calculi. A stone in the left renal calyx identified by conventional-dose computed tomography (CDCT; 120 kVp, 250 mAs; size: 3.2 mm×2.9 mm×2.5 mm) (A) and low-dose computed tomography (LDCT; 120 kVp, 60 mAs; size: 3.1 mm×2.9 mm×2.6 mm) (B). A stone in the right lower ureter identified by CDCT (size: 4.5 mm×6.3 mm×5.2 mm) (C) and LDCT (size: 4.4 mm×6.4 mm×5.3 mm) (D). The white arrows indicate the location of the stone.

4. Image Quality Evaluation for CDCT and LDCT

Following the reconstruction of CDCT and LDCT images, the image quality, noise, and diagnostic confidence were evaluated by radiologists. Based on their assessment, the image noise was comparable between the two cohorts, and no significant differences were observed in image quality (p>0.05) (Table 3).

Table 3

Comparison of Image Quality, Image Noise, and Diagnostic Confidence between CDCT and LDCT

5. Radiation Dose Comparison between CDCT and LDCT

The mean CTDI_vol, DLP, and ED of CDCT were 18.32± 3.60 mGy, 486.41±78.72 mGy·cm, and 5.94±1.34 mSv, respectively. The mean CTDI_vol, DLP, and ED of LDCT were 3.43±0.83 mGy, 101.33±65.41 mGy·cm, and 1.23±0.78 mSv, respectively. Compared to CDCT, the LDCT radiation dose was significantly reduced (p<0.05) (Table 4).

Table 4

Comparison of CT Radiation Dose between Modalities

Discussion

LDCT has been generally limited in clinical applications due to its poor image quality and high noise levels. In the present study, our in vitro model identified 60 mAs as the optimal dose, demonstrating the highest detection rate. Therefore, this dose was used in the clinical trials. Compared to CDCT, LDCT did not significantly affect clinical diagnosis outcomes, indicating that LDCT is feasible for use in clinical settings. Significant differences in CTDI_vol, DLP, and ED were observed between the two groups (p<0.05). Based on these results, although LDCT substantially reduced radiation exposure, the image quality and diagnostic accuracy were maintained. Together, these findings confirm the feasibility of LDCT in clinical applications. It should be noted; however, that this conclusion is specific to the evaluation of urinary calculi, as image quality may vary when assessing other pathologies.

Urinary calculi are commonly found in the kidneys, ureters, urinary bladder, and urinary tract [16], and they typically manifest as hematuria, flank pain, abdominal pain, nausea, and vomiting [17]. Urinary calculi can cause significant pain and lead to dysuria, which, if left untreated, may progress to anuria and uremia. Therefore, early detection is critical for improving patient outcomes. Unfortunately, traditional diagnostic methods have a relatively high missed diagnosis rate. While KUB is easy to perform, its diagnostic accuracy is relatively low (specificity 76%). US can preliminarily detect hydronephrosis in the urinary system, but it is less effective in detecting urinary calculi (specificity 53%). In contrast, CT significantly improves image resolution and diagnostic accuracy, with a specificity and sensitivity exceeding 95% [18]. However, with the widespread use of CT, radiation exposure has markedly increased. Therefore, it is essential to maintain high image quality while minimizing radiation dose.

Several studies [19–21] have shown that LDCT, compared to CDCT, provides similar stone detection rates while significantly reducing radiation exposure, with no significant difference in image quality. In clinical practice, instruments from various manufacturers, including GE, Siemens, and Philips, are used for LDCT-based detection of urinary calculi. However, the models and internal structures of these machines vary, meaning there is currently no unified standard for LDCT-based assessment.

The primary approaches to reducing radiation exposure in clinical practice include increasing pitch, reducing tube voltage, and lowering tube current [22–24]. Increasing pitch, however, increases the risk of missed small lesion. Reducing tube voltage can drastically degrade image quality. Lowering tube current enhances image noise, which may negatively affect the quality of low-contrast tissue image. The density and spatial resolution of the 64-slice spiral CT are satisfactory. Following imaging, the morphology of the urinary tract can be visualized from multiple angles and dimensions, which effectively compensates for any deficiencies in LDCT imaging resolution. In this study, we demonstrated that CDCT and LDCT produced consistent stone detection rates.

This study had several limitations. First, while preliminary findings from the in vitro model guided the selection of 60 mAs. Although 100% detection accuracy was achieved at 60 mAs, the sample size was relatively limited. Higher tube loading settings, such as 80 mAs or 100 mAs, may improve detection accuracy, especially for stones with lower calcium content or in more complex pathological scenarios. It is important to note that, while LDCT was effective for detecting urinary calculi, its utility in diagnosing other causes of symptoms, such as low back pain or hematuria, may be limited. These symptoms can also arise from conditions like renal tumors, pyelonephritis, or diseases affecting other abdominal organs, including the liver, gallbladder, pancreas, and bowel. Therefore, LDCT should be used in conjunction with other diagnostic methods for a comprehensive evaluation and accurate diagnosis. Additionally, the initial scanning was performed with a slice thickness of 5 mm, which may result in signal averaging and obscure finer details. Although the images were later reconstructed to a slice thickness of 0.625 mm using the FBP method, partial volume effects could still affect the clarity of smaller structures. Furthermore, as per our study design, patients who tested positive on CDCT were subsequently tested with LDCT, leading to additional radiation exposure. Second, the ex vivo pork kidney model used in this study does not fully replicate the human abdominal environment, particularly in terms of tissue density and X-ray absorption. While we used human urinary calculi of various compositions, we focused on determining the optimal low dose for detecting stones of different compositions and sizes, rather than optimizing the dose specifically for each stone composition. Third, we did not assess the impact of CT number fluctuations caused by stone composition or noise effects. This is particularly important because low doses may miss stones with lower calcium content, which are more challenging to detect due to their lower CT values. Given these limitations, further research should focus on optimizing the dose for different stone compositions, expanding the sample size for more robust conclusions, and investigating the impact of CT number fluctuations and noise on stone detection. Additionally, future studies could consider using in vivo models that more accurately mimic human tissue properties to allow for a more comprehensive evaluation of LDCT’s clinical efficacy.

Conclusion

In conclusion, LDCT substantially reduces patient radiation exposure while maintaining acceptable image quality, demonstrating its significant potential for clinical application.

Article information Continued