Ultra-low-dose computed tomography of the paranasal sinuses: evaluating diagnostic adequacy for chronic rhinosinusitis

Introduction

Computed tomography (CT) of the paranasal sinuses and facial bones is central to the diagnosis and delineation of sinus disease (1,2). CT provides complementary information to the history, physical examination and nasendoscopy, allowing assessment of mucosal disease and relevant anatomy. The concern, as with all CT imaging, is the degree of radiation exposure that accompanies this investigation. This is particularly relevant in paranasal sinus imaging due to the proximity of critical structures. The optic lens, which is susceptible to radiation induced cataract formation, is included in the scan volume, and exposed to both direct and scattered radiation (1,3). The thyroid gland, although outside the scan volume, also receives a detectable dose (4). Patients with inflammatory sinus disease frequently require repeat imaging to assess disease progression or treatment response, further increasing the risk of stochastic and deterministic effects of radiation exposure.

The paranasal sinuses have high intrinsic subject contrast given the interface of air, bone, and soft tissue. This contrast may allow for significant reductions in CT radiation dose while still maintaining accurate delineation of anatomy and pathology of the sinuses. Studies have employed various techniques to facilitate dose reduction including beam hardening by use of tin filtration, low kV techniques, and use of iterative reconstruction (5-9). Petritsch et al. achieved diagnostic image quality with an estimated dose of 0.012 mSv using tin filtration and iterative reconstruction (5). However, tin filtration is not available on all clinically implemented CT scanners. Cone beam CT has also been used for low dose paranasal sinus imaging (10); but requires the additional expense and resources of a dedicated scanner, which is not standard in many centres.

The purpose of this pilot study was to design an ultra-low-dose (ULD) CT sinuses technique that provided adequate image quality using conventional CT scanning equipment. Chronic rhinosinusitis (CRS) was used as the index condition, and our hypothesis was that ULD CT would be adequate to identify the presence of sinonasal mucosal thickening. The secondary aim was to assess the agreement in disease severity using the Lund-Mackay (LM) score, between the standard-dose (SD) and ULD protocol.

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Metro North Health Human Research Ethics Committee (project ID: HREC/2021/QRBW/78770), which approves research projects conducted through the Royal Brisbane and Women’s Hospital at the Herston campus, where the Ear, Nose, and Throat (ENT) department is located. Informed consent was taken from all individual participants before enrolment. The study is reported according to the STROBE reporting guidelines (available at https://www.theajo.com/article/view/10.21037/ajo-2025-1-74/rc).

A total of 11 participants were enrolled from the outpatient ENT clinic at the Royal Brisbane and Women’s Hospital. As a pilot study, no formal sample size calculation was performed, and participant numbers were determined pragmatically; accordingly, the study was not powered to detect statistically significant differences between groups. Participants below the age of 18 years and those who were pregnant were excluded from enrolment. All participants included had a prior diagnosis of CRS, confirmed through history and examination with an ENT surgeon. Scans were performed on a Siemens Biograph mCT Flow using the standalone CT mode (equivalent to the standard diagnostic CT counterpart Siemens Somatom Definition AS).

Image acquisition

Each participant underwent two scans: the vendor SD CT protocol available on the Siemens scanner, and the test ULD CT protocol. Eighty kV was selected as it significantly lowered the patient dose at the expense of potential beam hardening artefact. A fixed tube current of 20 mAs was used as the head does not vary significantly in the x-y plane between patients. Given the inherent intrinsic contrast within the sinuses is high, a higher degree of noise is tolerable. The Sinogram Affirmed Iterative Reconstruction (SAFIRE) protocol was utilised as it can reduce image noise as well as beam hardening artefact. Acquisition parameters are detailed in Table 1. Acquisitions were reconstructed with a slice thickness of 2 mm at increments of 2 mm. The ULD CT images underwent iterative reconstruction during post-processing with a sharp HR60 kernel applied to optimise bony detail. Soft tissue reconstructions were also provided using a BF37 kernel on the ULD images. Planar reconstruction in axial, sagittal, and coronal planes were provided to the four reporters.

Subjective image quality analysis

Four reporters independently reviewed the images: two consultant ENT surgeons and two radiologists (one senior registrar and one senior consultant). Reporters were blinded to participant details and acquisition protocol, and images were presented in a randomised order. Reporters first rated image quality subjectively on a 5-point Likert scale, with a rating of 1 indicating non-diagnostic and a rating of 5 indicating excellent image quality. Reporters then utilised the LM rating system described by Lund et. al. to grade the degree of sinus disease, with a maximum score of 24 (11). A difference of >2 points between SD and ULD datasets was defined as clinically significant.

Dosimetry

The dose length product (DLP) and volume CT dose index (CTDI_vol) were collected from the dose report of the studies performed. Absorbed dose to the lens, absorbed dose to the thyroid, and the effective doses were calculated using CT-Expo v2.5, a tool for dose evaluation in CT (12).

Statistical analysis

Inter-reporter reliability of the LM score, based on single ratings, was assessed for SD and ULD CT datasets separately using a two-way, random-effects model with absolute agreement (13,14). Inter-reporter agreement on the presence or absence of sinusitis across all four reporters was assessed using Fleiss’s kappa statistic, as was agreement on image quality ratings. Categorical variables were summarised as frequencies and percentages. Continuous variables were summarised as means with range or as medians with interquartile range (IQR) where non-normally distributed. Statistical analyses were performed in the R Statistical package (version 4.1.3).

Results

Images from all 11 patients scanned were deemed adequate for interpretation by the four reporters, with none excluded from analysis because of image quality. Image quality scores were higher for SD compared with ULD [median 4.5 (IQR, 4.5-4.75) vs. 3.5 (IQR, 3.0-3.5)] and were consistently rated equal or higher across all patients.

Dosimetry

The mean CTDI_vol and DLP for the SD CT scans were 8.05±1.12 mGy (range, 6.7–9.9 mGy) and 126.54±22.04 mGy·cm (range, 91–165.8 mGy·cm). For the ULD technique, the CTDI_vol was 0.87 mGy for all scans and the mean DLP was 12.85±1.04 mGy.cm (range, 11–14.4 mGy·cm). The absorbed dose to the optic lens ranged from 6.7–9.9 mGy for the SD examinations and was 0.8 mGy for ULD. The absorbed dose to the thyroid was 0.2–0.3 mGy for the SD examinations and calculated as 0 mGy for ULD. The effective dose was 0.2–0.3 mSv for the SD and calculated as 0 mSv for the ULD protocol.

LM score

LM scores for each patient/scan and reporter are available in Table S1. Inter-reporter reliability for LM scores for SD and ULD scans is displayed in Table 2. Interclass correlation coefficients (ICC) demonstrated excellent inter-rater reliability for both protocols: ICC =0.96 [95% confidence interval (CI): 0.89–0.99] for SD CT, and ICC =0.94 (95% CI: 0.86–0.98) for ULD CT.

Presence/absence of sinusitis

Kappa agreement for presence or absence of sinonasal mucosal thickening between SD CT and ULD CT for each reporter is displayed in Table 3. Agreement between doses across reporters was good to very good, with kappa scores ranging from 0.79 to 1.0.

Discussion

This pilot study demonstrates the utility of a ULD CT protocol in patients with sinus disease, using CRS as the index condition. In this small sample, the ULD CT protocol provided sufficient image quality for both for detecting the presence or absence of sinonasal mucosal disease, and for grading disease severity using the LM score. This highlights the potential of ULD CT as a screening and monitoring tool for sinusitis. As illustrated in Figure 1, our ULD CT protocol (Figure 1A) preserves sufficient image quality to visualise sinonasal mucosal thickening with only minimal loss of resolution compared with SD CT (Figure 1B). Given that patients with CRS often require repeat imaging, reducing cumulative radiation exposure represents a valuable safety measure.

Figure 1 Comparison between ultra-low-dose CT scan (A) and standard-dose CT scan (B) of the paranasal sinuses in participant with chronic rhinosinusitis. CT, computed tomography.

The ULD CT protocol achieved a radiation dose reduction of approximately 90% compared to the SD protocol, in line with the radiologic principle of keeping exposure “as low as reasonably achievable” (ALARA). These findings are consistent with previous studies showing that substantial dose reduction in paranasal sinus CT can be achieved without compromising diagnostic accuracy, through specialised techniques (5,6,9). Although our protocol did not reach the extremely low dose levels reported with advanced scanners employing tin pre-filtration (5,9), we achieved comparable or lower doses than other ULD CT studies utilising tube voltage and current reduction with iterative reconstruction (6,15,16). Further, by using a standard CT scanner configuration, our protocol improves feasibility for routine clinical use and broadens accessibility.

Variability in radiation dose was greater with the SD CT. The vendor SD scan protocol used mAs modulation and variable kV settings. Despite the facial bones and sinuses being relatively homogenous across the population, there was a wide range in the delivered CTDI_vol and DLP. This variability highlights the influence of vendor-specific dose modulation algorithms and suggests that SD protocols may deliver higher doses than necessary. In contrast, the ULD CT protocol provided more consistent and predictable dose exposure.

In some ULD CT reconstructions, a loss of fine bony detail was observed. For example, Figure 2 illustrates an apparent dehiscence of the lamina papyracea on ULD CT (Figure 2A) that is not seen on the SD CT scan (Figure 2B). Accurate assessment of fine osseous structures was not the primary objective of this study, which focussed on detecting and grading sinonasal mucosal disease. Evaluation of other sinonasal pathologies requiring detailed anatomical assessment—such as cerebrospinal fluid leaks, tumours, or orbital and intracranial disease extension—was also outside the scope of this study. Accordingly, this ULD CT protocol should not be used as a standalone imaging modality for these pathologies or for surgical planning (17). In patients proceeding to operative management, confirmatory SD CT could be performed with minimal incremental radiation exposure. Achieving an ULD CT of sufficient quality for surgical planning remains a potential target for future research and has been reported by others. Diklić et al. (18) found that sufficient image quality for functional endoscopic sinus surgery planning could be achieved with a ULD CT protocol at a slightly higher dose of CTDI_vol 2.2 mGy, DLP 22.9 mGy.cm, approximately double the dose achieved in our protocol.

Figure 2 Comparison between ultra-low-dose CT scan (A) and standard-dose CT scan (B) of the paranasal sinuses, demonstrating loss of bony detail on lamina papyracea (green arrows). CT, computed tomography.

While the study demonstrates promising results, our findings must be interpreted in the context of several limitations. First, the sample size was small, and recruited from a single tertiary centre, limiting the statistical power and generalisability of the results. Second, although reporters were blinded to acquisition protocol, the difference in image noise and reconstruction characteristics between the ULD and SD CT may have allowed reporters to infer the scan type, introducing potential observer bias. Finally, these findings are specific to the scanner model and reconstruction algorithms used in this study, and performance may differ across other CT platforms. Nonetheless, assessment of sinonasal mucosal disease relies on anatomical structures with high intrinsic contrast between air, soft tissue, and bone. This large separation in Hounsfield units means that the diagnostic task is relatively robust to increased image noise, suggesting that similar ULD protocols may be achievable on other modern CT systems.

Our ULD CT protocol may therefore have potential as a screening modality for patients in whom the presence or absence of sinonasal mucosal disease is uncertain. Potential applications include acute presentations such as periorbital cellulitis, and elective scenarios such as the investigation of facial pain or chronic cough. There is also a benefit in reduced cumulative radiation exposure for CRS patients who often receive repeated scans to monitor disease progression and response to treatment. Utilising widely available CT equipment, our technique has the potential to improve access to ULD CT of the sinuses to a wider community, including regional and rural areas. Future research should validate these findings in larger, multi-centre cohorts and explore the performance of ULD CT across a wider spectrum of sinonasal disease and clinical situations including disease relapse after surgery, paediatric populations and operative navigation.

Conclusions

This pilot study evaluated an ULD CT protocol for the detection of sinonasal mucosal disease using conventional CT scanning equipment. These preliminary findings suggest ULD CT can provide sufficient image quality to identify and grade sinus disease, at a substantially lower radiation dose compared with SD CT. ULD CT may therefore represent a practical and accessible tool for the screening or monitoring of sinusitis; however, larger studies are required to confirm diagnostic performance and define its role beyond mucosal disease assessment.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/ajo-2025-1-74/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-2025-1-74/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-2025-1-74/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-2025-1-74/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Metro North Health Human Research Ethics Committee (project ID: HREC/2021/QRBW/78770), which approves research projects conducted through the Royal Brisbane and Women’s Hospital at the Herston campus, where the Ear, Nose, and Throat (ENT) department is located. Informed consent was taken from all individual participants before enrolment.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Zammit-Maempel I, Chadwick CL, Willis SP. Radiation dose to the lens of eye and thyroid gland in paranasal sinus multislice CT. Br J Radiol 2003;76:418-20. [Crossref] [PubMed]
2. Riechelmann HEuropäischen Akademie für Allergie und Klinische Immunologie (EAACI) und der European Rhinologic Society (ERS). Chronic Rhinosinusitis - EPOS 2012 Part I. Laryngorhinootologie 2013;92:193-201; quiz 202-3. [Article in German]. [Crossref] [PubMed]
3. Czechowski J, Janeczek J, Kelly G, et al. Radiation dose to the lens in sequential and spiral CT of the facial bones and sinuses. Eur Radiol 2001;11:711-3. [Crossref] [PubMed]
4. Schmale IL, Vandelaar LJ, Luong AU, et al. Image-Guided Surgery and Intraoperative Imaging in Rhinology: Clinical Update and Current State of the Art. Ear Nose Throat J 2021;100:NP475-86. [Crossref] [PubMed]
5. Petritsch B, Kosmala A, Weng AM, et al. Tin-filtered 100 kV ultra-low-dose CT of the paranasal sinus: Initial clinical results. PLoS One 2019;14:e0216295. [Crossref] [PubMed]
6. Bang M, Choi SH, Park J, et al. Radiation Dose Reduction in Paranasal Sinus CT: With Feasibility of Iterative Reconstruction Technique. Otolaryngol Head Neck Surg 2016;155:982-7. [Crossref] [PubMed]
7. Rosenfeld RM, Piccirillo JF, Chandrasekhar SS, et al. Clinical practice guideline (update): adult sinusitis. Otolaryngol Head Neck Surg 2015;152:S1-S39. [Crossref] [PubMed]
8. Prulière-Escabasse V, Coste A. Image-guided sinus surgery. Eur Ann Otorhinolaryngol Head Neck Dis 2010;127:33-9. [Crossref] [PubMed]
9. Grunz JP, Petritsch B, Luetkens KS, et al. Ultra-Low-Dose Photon-Counting CT Imaging of the Paranasal Sinus With Tin Prefiltration: How Low Can We Go?. Invest Radiol 2022;57:728-33. [Crossref] [PubMed]
10. Al Abduwani J, ZilinSkiene L, Colley S, et al. Cone beam CT paranasal sinuses versus standard multidetector and low dose multidetector CT studies. Am J Otolaryngol 2016;37:59-64. [Crossref] [PubMed]
11. Lund VJ, Mackay IS. Staging in rhinosinusitus. Rhinology 1993;31:183-4. [PubMed]
12. Stamm G, Nagel H. CT-Expo v2.5 A Tool for Dose Evaluation in Computed Tomography [Software]. Science & Technology for Radiology; 2017. Available online: http://www.sascrad.com/information/downloads/.
13. Caldwell AR. SimplyAgree: An R package and jamovi Module for Simplifying Agreement and Reliability Analyses. J Open Source Softw 2022;7:4148.
14. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135-60. [Crossref] [PubMed]
15. Bulla S, Blanke P, Hassepass F, et al. Reducing the radiation dose for low-dose CT of the paranasal sinuses using iterative reconstruction: feasibility and image quality. Eur J Radiol 2012;81:2246-50. [Crossref] [PubMed]
16. Bodelle B, Wichmann JL, Klotz N, et al. Seventy kilovolt ultra-low dose CT of the paranasal sinus: first clinical results. Clin Radiol 2015;70:711-5. [Crossref] [PubMed]
17. O'Brien WT Sr, Hamelin S, Weitzel EK. The Preoperative Sinus CT: Avoiding a "CLOSE" Call with Surgical Complications. Radiology 2016;281:10-21. [Crossref] [PubMed]
18. Diklić A, Valković Zujić P, Šegota D, et al. Optimization of paranasal sinus CT procedure: Ultra-low dose CT as a roadmap for pre-functional endoscopic sinus surgery. Phys Med 2020;78:195-200. [Crossref] [PubMed]