F1000Research F1000Research 2046-1402 F1000 Research Limited London, UK 10.12688/f1000research.110971.1 Research Article Articles Low dose protocol for high resolution CT thorax: influence of matrix size and tube voltage on image quality and radiation dose

[version 1; peer review: 1 approved, 1 approved with reservations, 1 not approved]

 Kumar Navish Data Curation Methodology Writing – Original Draft Preparation 1 Pradhan Abhimanyu Data Curation Writing – Review & Editing https://orcid.org/0000-0002-2910-5338 1 Kadavigere Rajagopal Conceptualization Formal Analysis Investigation https://orcid.org/0000-0003-3486-8740 2 Sukumar Suresh Investigation Methodology https://orcid.org/0000-0001-9345-9790 a 1 Department Medical Imaging Technology, Manipal College of Health Professions, Manipal Academy of Higher Education, Manipal, Manipal, Karnataka, 576104, India Department of Radio-diagnosis & Imaging, Kasturba Medical College, Manipal Academy of Higher Education, Manipal, Karnataka, 576104, India suresh.sugumar@manipal.edu

No competing interests were disclosed.

 6 4 2022 2022 11 399 29 3 2022 Copyright: © 2022 Kumar N et al. 2022 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background: High-resolution CT (HRCT) thorax has increase demand due to its advantage in diagnosing chronic respiratory diseases. The feasibility of matrix size with different tube voltage in the HRCT protocol of thorax is unknown. Therefore, this study aimed to compare the effect of matrix sizes and tube voltage on image quality and radiation dose on adult HRCT thorax.

Methods: A Phantom experiment was performed, followed by a patient scan. For phantom and patient scan, a total of six protocols with two tube voltage settings, 120 kVp and 100 kVp, with a combination of three matrix sizes, 512, 768, and 1024 were used. In this study, 180 adult patients who had HRCT thorax scan were considered. Dose data was collected, and quantitative image analysis was performed by drawing region of interests on the acquired phantom and patient images. Qualitative image analysis was performed independently by two blinded radiologists.

Results: The dose report of the phantom experiment revealed that the 100kVp with selected matrix size delivered 15.64% and 15.62% less radiation dose in terms of volumetric computed tomography dose index (CTDIvol) and dose length product (DLP), respectively, compared to 120kVp settings with selected matrix sizes. Similarly, for the patient population, the CTDIvol and DLP difference noted for 120kVp and 100kVp with different matrix sizes was statistically significant (p<0.001). For quantitative image quality, the difference noted was also statistically significant among two kVp settings. The mean score for subjective image assessment was greater than 4.5 for diagnostic acceptability and streak artefacts.

Conclusion: The result suggests that the 100 kVp with 512 X 512 matrix size is preferable in the HRCT Lung to achieve the optimal diagnostic image quality with a reduction of almost 40% of the dose to the patients compared to 120 kVp techniques.

 HRCT thorax low dose dose reduction matrix size tube voltage Manipal Academy of Higher Education Manipal Academy of Higher Education, Manipal, Karnataka. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Introduction

The prevalence of respiratory diseases in India is substantially high as per the recent global burden of disease survey data ( Dandona et al., 2017). Chronic respiratory diseases, especially asthma and obstructive pulmonary disease, are of particular importance for having wide variations in morbidity and mortality in different Indian states ( Salvi et al., 2018). Additionally, the novel COVID-19 outbreak declared by the World health organization (WHO) has increased the burden of respiratory disease in India as well as globally ( Cucinotta & Vanelli, 2020; To et al., 2020). The outbreak of COVID-19 led to an increase in the demand for image modalities for the diagnosis of the disease. A chest radiograph plays a crucial role in diagnosing suspected respiratory diseases. High resolution computed tomography (HRCT) of the thorax is one such examination in high demand in routine use and for initial and rapid diagnosis of COVID-19 pneumonia with the low rate of missed diagnosis ( Kashyape & Jain, 2021; Li & Xia, 2020). Although reverse-transcription polymerase chain reaction (RT-PCR) is the preferred tool for diagnosing novel COVID-19 infection, HRCT thorax has better sensitivity ( Kalra, Homayounieh, Arru, Holmberg & Vassileva, 2020). The benefit of HRCT is not only limited to disease confirmation, but it plays a significant role in determining disease progression, treatment response, management, and decision on hospitalization and discharge of the COVID-19 patient through subsequent scans performed during these periods ( Kalra et al., 2020; Rubin & Ryerson & Haramati, 2020; Azadbakht et al., 2021; Davarpanah et al., 2020).

The use of imaging modalities during the COVID-19 pandemic indeed has maximum benefit, but the fact that radiation exposure increases the risk of cancer should not be ignored. The linear-no threshold (LNT) model describes cancer risk even with a low dose of radiation ( Shah, Sachs & Wilson, 2012; Mullenders, Atkinson, Paretzke, Sabatier & Bouffler, 2009). The risk of radiation-induced cancer during the pandemic is of concern, primarily due to the increase in COVID-19 infection and the need for multiple scans required during the recovery period of the disease ( Wang, Liu, Yang & Chen, 2020). As recommended by International Commission on Radiological Protection (ICRP), the radiation dose delivered to the patient should be as low as reasonable achievable (ALARA) even during the pandemic, and the dose optimization for COVID-19 patients is necessary as there is no low dose HRCT protocol reported ( Azadbakht et al., 2021; Yeung, 2019). There are several techniques to optimize the radiation dose for HRCT, such as reducing tube voltage, tube current, scan length, high pitch techniques, automatic exposure control (AEC), and iterative reconstruction techniques ( Azadbakht et al., 2021). The feasibility of matrix size with different tube voltage in the HRCT protocol of thorax is unknown. Therefore, this study aimed to compare the effect of matrix sizes and tube voltage on image quality and radiation dose on adult HRCT thorax. The working hypothesis of this study was that the change in tube voltage could influence the image quality.

 Methods

The study approval was obtained from the institutional ethics committee (IEC: 168/2019), Kasturba hospital, Manipal. The study was carried out in department of Radio-diagnosis and Imaging, Kasturba Hospital, Manipal. An experimental study was initially tested on Phantom to understand the influence of tube voltage and matrix size on the quality of the image and dose. The outcome of the results was then tested on the patient’s population.

Phantom study

This experiment was conducted in a 32 cm body CT polymethyl metacrylate (PMMA) phantom (Unfors, serial no. 0143) using an iCT (Philips Medical Systems) 128-row multiple slice CT scanner. The phantom was positioned on the CT table. The height of the phantom was adjusted using the position of laser light. The experiment was performed at two different tube voltage settings (120 kVp and 100 kVp) and three matrix sizes (512, 768 & 1024) available in the CT scanner ( Table 1). All the other parameters, Reconstruction mode iDose4 level 2, collimation 64*0.625, gantry tilt of 0 degrees, the pitch of 1, rotation time: 0.5, slice thickness: 1 mm, reconstruction increment of 0.7500, a field of view (FOV) of 350 mm, and the adaptive filter were kept constant. The automatic tube current modulation was enabled during the procedure. An axial image with 1 mm thickness was obtained. Each protocol was scanned with a 130 mm scan length. Phantom CT image obtained using different parameters was analysed for dose and quantitative image quality. Signal-to-noise (SNR) and contrast-to-noise (CNR) ratios were analysed by drawing five regions-of-interest (ROIs) of 100 mm 2. Out of five ROIs, four were positioned in peripheral location (3, 6, 9, 12 o’clock) and one in the centre location of the phantom along with one ROI in the background to encompass homogeneity of measured regions ( Figure 1).

Protocol used to scan phantom and patient with various combination of kVP and matrix size.
Protocol Tube voltage settings Matrix size
A 120 512 × 512
B 120 768 × 768
C 120 1024 × 1024
D 100 512 × 512
E 100 768 × 768
F 100 1024 × 1024
Location of ROIs on CT PMMA phantom. Study population

A total of 180 patients (Mean age 54.50 ±15.81, Male 55.06 ±17.15, Female 53.85±14.17) who were 18 years and above and prescribed for the HRCT Thorax were considered for this study (See Underlying data) ( Sukumar, 2022). The written informed consent was obtained from the participant. All the scan was performed in iCT (Philips Medical Systems) 128-row multiple slice CT scanner with six different scanning protocol similar to phantom ( Table 1). The automatic tube current modulation (Dose Right Index: 17) was enabled during the procedure.

After acquiring the image using a different protocol, the subjective image assessment of the patient’s population was performed according to the European Guidelines on Quality Criteria for CT ( Menzel et al., 2000). Two radiologists with a mean experience of 12.00 ± 1.41 years, reviewed images independently. The radiologists were blinded to the scanning protocols and the procedure. Images were displayed in the HRCT window (-600 WL, 1600 WL) on an Advantage Workstation (Philips Healthcare). Radiologists’ evaluation for the diagnostically acceptable image quality for HRCT thorax was considered (See Underlying data) ( Sukumar, 2022). The diagnostic acceptability of the image quality was rated on five-point rating scale were, 1 = poor for diagnosis, 2 = suboptimal for diagnosis, 3 = good for diagnosis, 4 = very good for diagnosis, 5 = excellently. The streak artefacts seen anywhere on the CT image was rated on a five-point rating scale, were 1= artefact causing diagnosis impossible, 2 = artefacts affecting diagnostic information, 3 = major artefacts affecting visualization of major structures but diagnosis still possible, 4 = minor artefacts not interfering with diagnostic decision-making, and 5 = no artefacts.

The quantitative image quality assessment was assessed by drawing 100 mm 2 ROIs on high density structures of thorax, such as subscapularis muscle, ascending aorta, and descending aorta (See Underlying data) ( Sukumar, 2022) ( Figure 2).

Location of ROIs on (a) subscapularis muscle, (b) ascending aorta, (c) descending aorta and (d) background for quantitative image quality assessment in HRCT thorax.

The SNR, CNR, and Figure of Merit (FOM) ratios for the phantom and the participant population were analysed by the following equation ( Chang, Hsu, Lin & Hsu, 2017) (See Underlying data) ( Sukumar, 2022). SNR = CT ROI / SD ROI CNR = CT ROI – CT Background / SD Background FOM = CNR 2 / CTDI vol

Where CT ROI and CT Background are the average Hounsfield unit of the corresponding ROI, and SD ROI and SD Background are the standard deviations of the corresponding ROI.

The dose data (CTDI vol and DLP) for both phantom study and patient population study were collected from the dose info of the system. For the study population the obtained DLP was further used to calculate the effective dose by multiplying with a conversion factor of 0.014 ( Deak, Smal & Kalender, 2010) (See Underlying data) ( Sukumar, 2022).

 Statistical analysis

Statistical Package for the Social Sciences (SPSS) Statistics for Windows, Version 16 ( Kumar, Pai & Vineetha, 2020) was used to perform the statistical analysis. For the study population the obtained data were not normally distributed. Therefore, the median and quartiles are reported, and for comparison of quantitative image analysis and dose matrix, Mann Whitney U ( Kumar, Pai & Vineetha, 2020) test was performed. Mean score for each protocol was calculated for subjective image assessment. P<0.05 was considered statistically significant.

 Results Quantitative image analysis of phantom image

The influence of various combinations of tube voltage (120 and 100 kVp) and matrix size (512, 768 and 1024) on image quality (SNR and CNR) and dose matrix (CTDI vol and DLP) are summarized in Table 2. The difference noted (in percentage) when comparing different protocols are outlined in Table 3. In this study, the phantom was scanned with two tube voltage techniques combined with three different matrix sizes (a total of six scanning protocols). When comparing the image quality of 120 kVp with various matrix sizes, protocol A showed the highest values compared to protocols B and C. The dose changes noted were very minimal.

Comparison of 120 kVp and 100kVp on image matrix size of 512, 768 and 1024: Phantom experiment.
Protocol A B C D E F
Tube voltage 120 100
Matrix size (Mean) 512 768 1024 512 768 1024
SNR (Mean) 0.91 0.70 0.74 0.84 0.60 0.55
CNR (Mean) 21.15 8.75 9.92 16.65 10.09 9.07
CTDI vol (mGy) (Mean) 23.22 23.22 22.11 19.85 19.85 19.85
DLP (mGy *cm) (Mean) 346.18 346.18 329.20 296.00 296.00 296.00
Table showing the percentage of difference in SNR, CNR, CTDI and DLP when comparing with different groups of protocol.
Protocol comparison Difference in %
SNR CNR CTDI DLP
A vs B 26.08 82.94 0 0
A vs C 20.60 72.28 4.89 5.02
A vs D 8 23.80 15.64 15.62
A vs E 41.05 70.80 15.64 15.62
A vs F 49.31 79.94 15.64 15.62
B vs C 5.55 12.53 4.89 5.02
B vs D 18.18 62.65 15.64 15.62
B vs E 15.38 14.22 15.64 15.62
B vs F 24 3.59 15.64 15.62
C vs D 12.65 50.65 15.64 10.62
C vs E 20.89 1.69 15.64 10.62
C vs F 29.45 8.95 15.64 10.62
D vs E 33.33 49.06 0 0
D vs F 41.72 58.94 0 0
E vs F 8.69 10.09 0 0

Similarly, when comparing the image quality of 100 kVp with various matrix sizes, protocol D showed higher values than protocol E and protocol F, and no changes were noted in terms of dose matrix ( Table 2). The result of the phantom study showed that, as the matrix size increases, the SNR and CNR decreases, except for the case of protocol C, where the SNR and CNR values were slightly higher than protocol B. Therefore, when comparing the image quality of the 120 kVp technique and 100 kVp technique with various matrix sizes, the matrix size of 512 (protocol A and protocol D) produced better image quality than another matrix size. When comparing among two tube voltages with the same matrix size of 512 (protocol A and protocol D), protocol A showed better result than protocol D when considering image quality where the difference (in percentage) among these two protocols for SNR and CNR was 8% and 23.80% respectively, which is minimal as compared to the difference of other protocol ( Table 3). However, the difference noted in CTDI vol and DLP among these two protocols was 15.64% and 15.62%, respectively ( Table 3), which shows that 100 kVp techniques have reduced radiation dose.

 Quantitative image analysis of patient population

Similar to the phantom study, the influence of tube voltage and matrix size on image quality and radiation dose was performed in the patient population. The collected patient data were not normally distributed. Hence the median and quartiles for image quality (SNR and CNR), radiation dose matrix (CTDI vol and DLP), and Figure of merit (FOM) are summarized in Table 4. The statistically significant value (p<0.05) for comparison of median among various protocols are outlined in Table 5.

Comparison of 120 kVp and 100kVp on image matrix size of 512, 768 and 1024: patient’s population.
Protocol A B C D E F
Tube voltage 120 kVp 100 kVp
Matrix 512 768 1024 512 768 1024
Age mean 55.16±18.08 56.20±10.63 53.03±13.68 57.40±13.97 47.10±18.24 58.13±17.42
mA mean 112.50±2.09 115.43±9.18 113.80±5.08 122.43±11.49 126.70±13.60 124.96±12.82
SNR Median (Q1,Q3) 0.99 (0.76, 1.14) 0.64 (0.53, 0.77) 0.63 (0.49, 0.78) 0.63 (0.51, 0.74) 0.44 (0.38, 0.55) 0.46 (0.35, 0.61)
CNR Median (Q1,Q3) 49.99 (42.47, 61.42) 38.21 (29.85,42.83) 30.83 (23.57,36.60) 32.37 (26.94, 41.38) 25.03 (20.99, 29.37) 23.55 (20.08, 27.82)
FOM Median (Q1,Q3) 274.06 (196.21,398.74) 144.39 (93.49,202.06) 103.56 (60.84, 146.88) 189.34 (119.07, 291.3) 97.50 (67.80, 151.52) 95.43 (64.31, 134.31)
CTDI (mGy) Median (Q1,Q3) 9.15 (9.07, 39.24) 9.15 (9.08, 9.30) 9.16 (9.08, 9.32) 5.68 (5.49, 6.20) 5.99 (5.60, 6.53) 5.90 (5.56, 6.53)
DLP (mGy-cm) Median (Q1,Q3) 362.53 (340.80, 384.05) 352.02 (332.15, 372.25) 372.75 (355.52, 387.92) 223.27 (198.56, 247.48) 237.39 (213.75, 272.34) 234.58 (220.01, 254.48)
Effective dose (mSv) 5.07 4.92 5.21 3.12 3.32 3.28
Table showing the p-values of quantitative image quality, dose matrix and figure of merit when comparing with different groups of protocol.
Protocols comparison p-value
CTDI DLP SNR CNR FOM
A vs B 0.567 0.196 <0.001 * <0.001 * <0.001 *
A vs C 0.267 0.395 <0.001 * <0.001 * <0.001 *
A vs D <0.001 * <0.001 * <0.001 * <0.001 * 0.003 *
A vs E <0.001 * <0.001 * <0.001 * <0.001 * <0.001 *
A vs F <0.001 * <0.001 * <0.001 * <0.001 * <0.001 *
B vs C 0.756 0.030 * 0.773 <0.012 * 0.034 *
B vs D <0.001 * <0.001 * 0.641 0.255 0.067
B vs E <0.001 * <0.001 * <0.001 * <0.001 * 0.033 *
B vs F <0.001 * <0.001 * <0.001 * <0.001 * 0.038 *
C vs D <0.001 * <0.001 * 0.953 0.255 <0.001 *
C vs E <0.001 * <0.001 * <0.001 * 0.007 * 0.906
C vs F <0.001 * <0.001 * 0.001 * 0.004 * 0.796
D vs E 0.110 0.156 <0.001 * <0.001 * <0.001 *
D vs F 0.387 0.171 <0.001 * <0.001 * <0.001 *
E vs F 0.549 0.756 0.976 0.745 0.859

Statistical significant (p<0.05).

When comparing the image quality of 120 kVp technique (standard dose) with three different matrix size, protocol A showed highest value when compared with protocol B and protocol C. The difference noted was statistically significant between protocol A and protocol B, and protocol A and protocol C but for protocol B and protocol C, difference in SNR was not statistically significance whereas CNR was statistically significance. There was no difference noted for radiation dose when comparing within the protocols, except for the protocol B and protocol C where DLP showed statistical significance difference and FOM showed significant difference for all the comparisons ( Table 5). For 100 kVp technique (low dose), the result was in line with 120 kVp techniques, that is protocol D showed highest value in terms of image quality when compared with protocol E and protocol F, and the difference noted were statistically significant for all the comparison except for protocol E and protocol F. Similarly, there were no statistical difference noted in radiation dose matrix and FOM for all the comparisons except for comparison between protocol E and protocol F ( Table 5). The overall result showed that, with the increase in matrix size, the SNR, CNR, and FOM decrease for both standard 120 kVp and a low dose of 100 kVp techniques ( Figure 3).

Effect of tube voltage and matrix size on (a) SNR, (b) CNR, and (c) FOM.

The image quality for 512 × 512 matrix size with tube voltage settings of 120 kVp showed superior quality as compared to other combinations of tube voltage and matrix size, but when radiation dose was considered, the dose for 120 kVp with 512 matrix size (protocol A) was significantly higher than 100 kVp with 512 matrix size (protocol D) and other combinations ( Table 5). Nevertheless, the mean score for the subjective image assessment did not indicate a preference for any of the protocols in terms of diagnostic acceptability or streak artefacts ( Table 6).

Table showing the mean score for subjective image assessment.
Protocol Mean score
Diagnostic acceptability Streak artefacts
A 4.86 ± 0.22 4.78 ± 0.25
B 4.95 ± 0.15 4.93 ± 0.17
C 4.76 ± 0.28 4.65 ± 0.32
D 4.65 ± 0.35 4.60 ± 0.35
E 4.75 ± 0.34 4.63 ± 0.36
F 4.61 ± 0.33 4.60 ± 0.33
Discussion

The radiation dose in the HRCT thorax is high compared to the conventional lung imaging and exceeds 100 times the radiation dose of the Chest X-ray ( Ambrosino, Genieser, Roche, Kaul & Lawrence, 1994). Hence, optimising radiation dose and diagnostic accuracy of obtaining HRCT images is essential. The result of the present study state that with the use of a low dose, that is 100 kVp with the combination of 512×512 matrix size, the radiation dose to the patient can be reduced without compromising image quality.

The present study's findings on the effect of matrix size were limited to quantitative image quality, and no significant difference was noted on radiation dose. The result of the study revealed that the increase in matrix size decreased the SNR and CNR values, leading to noisier images for higher matrix size images. Similarly, Hata et al. (2018) reported the increase in noise and streak artefacts with larger matrix size, which reduces the application of larger matrix size in CT. Although the matrix size influences the spatial resolution of the image, increasing the matrix size seldom makes the spatial resolution better since the system's spatial resolution usually depends on the focal spot size of the X-ray tube and the size of the detector elements ( Hata et al., 2018). Also, the data size of the acquired images is larger with high matrix size images. Therefore, using a smaller matrix size, the images of HRCT thorax can be achieved without compromising spatial resolution.

In this study, the influence of tube voltage on image quality and radiation dose was analysed and revealed that using a low tube voltage of 100 kVp significantly reduced the radiation dose compared to 120 kVp. When comparing the effect of tube voltage along with matrix size, the study revealed that low tube voltage with low matrix size could achieve images without compromising diagnostic confidence. The institute protocol for HRCT thorax uses 120 kVp with 728 × 728 matrix size as a standard protocol. Compared with a low dose protocol of 100 kVp with a lower matrix size of 512 × 512, a significant reduction in radiation dose was achieved with no significant changes noted in quantitative image quality (Protocol B and D, Table 5). However, reducing the tube voltage below 100 kVp can be beneficial if it is tailored with patient weight, as reported by Desmet J. et al. (2021), where they used 80kVp for the patient weighing below 50 kg.

The effect of tube current on image quality and radiation dose was not performed in the present study as the protocol make use of automatic tube current modulation, which modulates the tube current according to the patient size. Although the radiation dose decreases linearly with the reduction of tube current but has limited use in HRCT as reducing tube current reduces the resolution of the images ( Azadbakht et al., 2021; Karabulut, Törü, Gelebek, Gülsün & Ariyürek, 2002). However, this can be overcome with the recent advancement in the reconstruction techniques, where filtered back projection (FBP) is being widely replaced by iterative reconstruction (IR) ( Honda et al., 2011). Literature reports that the use of low tube current with the combination of IR, such as Model-based Iterative Reconstruction (MBIR) and Adaptive Statistical Iterative Reconstruction (ASIR), can improve the compromised image quality due to the result of reduced tube current ( Yanagawa et al., 2014; Katsura et al., 2012; Tsukada et al., 2016). However, selecting the lowest possible tube current settings has benefited only over the selective examination, and it decreases the detectability of low-contrast details ( Karabulut et al., 2002; Neroladaki, Botsikas, Boudabbous, Becker & Montet, 2013). Therefore, it is preferred to use automatic tube current modulation for HRCT thorax.

The low dose HRCT thorax protocol with 100 kVp and 512 × 512 matrix can be recommended for COVID-19 patients. However, according to the webinar by International Atomic Energy Agency (IAEA) on CT practices and dose optimization, it was suggested to use HRCT thorax to diagnose COVID-19 only when the availability of RT-PCR is limited ( Kalra et al., 2020). Nevertheless, the use of HRCT thorax has several benefits when assessing disease progression and management of COVID-19 patients ( Davarpanah et al., 2020; Rubin et al., 2020). Also, there is an increasing demand for repeat scans for the patient with comorbidities, for the recovered individuals in evaluating pulmonary complications, and for research purposes ( Mahdavi et al., 2020).

The current study has a few Limitations. The present study was focused on SNR, CNR, and FOM. The correlation between the dose and quality of the image based on Body Mass Index (BMI) and chest circumference was not reported. The overall subjective diagnostic acceptability of the lung was reported. The visibility of the nodule, Broncho vascular bundle thickening, interlobular reticulation, and bronchiectasis was not reported individually.

 Conclusion

This study showed that 120 kVp with 512 × 512 matrix size produces superior image quality than 100 kVp techniques. However, the tube voltage of 100 kVp with the matrix size of 512 × 512 reduced almost 40% of radiation dose compared to 120 kVp techniques. Therefore, 100 kVp with 512 × 512 matrix size can be preferred for HRCT adult lung to achieve the optimal diagnostic image quality. This low dose protocol may be well adapted for the patients requiring routine HRCT and follow-up examination.

 Data availability Underlying data

Harvard Dataverse: Low dose protocol for high resolution CT thorax: influence of matrix size and tube voltage on image quality and radiation dose.

https://doi.org/10.7910/DVN/P3BL0F ( Sukumar, 2022)

This project contains the following underlying data:

que 1(B). (Ratings for diagnostic acceptability of the image quality)

que 1(E1-R2). (Combined ratings of two radiologists for diagnostic acceptability of the image quality)

que 2(B). (Ratings for streak artefacts seen on image)

que 2(R1-R2). (Combined ratings of two radiologists for streak artefacts seen on image)

ROI. (Quantitative values)

SNR, CNR. (Demographic details, radiation dose value and SNR and CNR values)

Data are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).

 Author contributions

All the authors have substantially contributed to the conception or design of the work. Data collection: Navish Kumar, Data analysis and Interpretation: Rajagopal Kadavigere and Abhimanyu Pradhan, drafting of article: Abhimanyu Pradhan and Suresh Sukumar, Critical Revision of the article: Suresh Sukumar and Rajagopal Kadavigere, Final Approval of the version to be published: Rajagopal Kadavigere.

 References Ambrosino MM Genieser NB Roche KJ : Feasibility of high-resolution, low-dose chest CT in evaluating the pediatric chest. Pediatr. Radiol. 1994;24(1):6–10. 8008501 10.1007/BF02017649 Azadbakht J Khoramian D Lajevardi ZS : A review on chest CT scanning parameters implemented in COVID-19 patients: bringing low-dose CT protocols into play. Egypt. J. Radiol. Nucl. Med. 2021;52(1):1–10. 10.1186/s43055-020-00400-1 Chang KP Hsu TK Lin WT : Optimization of dose and image quality in adult and pediatric computed tomography scans. Radiat. Phys. Chem. 2017;140(January):260–265. 10.1016/j.radphyschem.2017.02.006 Cucinotta D Vanelli M : WHO declares COVID-19 a pandemic. Acta. Biomedica. 2020;91(1):157–160. 32191675 10.23750/abm.v91i1.9397 Dandona L Dandona R Kumar GA : Nations within a nation: variations in epidemiological transition across the states of India, 1990–2016 in the Global Burden of Disease Study. Lancet. 2017;390(10111):2437–2460. 29150201 10.1016/S0140-6736(17)32804-0 Davarpanah AH Mahdavi A Sabri A : Novel Screening and Triage Strategy in Iran During Deadly Coronavirus Disease 2019 (COVID-19) Epidemic: Value of Humanitarian Teleconsultation Service. J. Am. Coll. Radiol. 2020;17(6):734–738. 32208138 10.1016/j.jacr.2020.03.015 Deak PD Smal Y Kalender WA : Multisection CT Protocols: Sex- and Age-specifi c Conversion Dose from Dose-Length Product 1 Purpose: Methods: Results: Conclusion. Radiology. 2010;257(1):158–166. 20851940 10.1148/radiol.10100047/-/DC1 Desmet J Biebaû C De Wever W : Performance of low-dose chest CT as a triage tool for suspected COVID-19 patients. J. Belg. Soc. Radiol. 2021;105(1):1–8. 33634237 10.5334/jbsr.2319 Hata A Yanagawa M Honda O : Effect of Matrix Size on the Image Quality of Ultra-high-resolution CT of the Lung: Comparison of 512 × 512, 1024 × 1024, and 2048 × 2048. Acad. Radiol. 2018;25(7):869–876. 29373211 10.1016/j.acra.2017.11.017 Honda O Yanagawa M Inoue A : Image quality of multiplanar reconstruction of pulmonary CT scans using adaptive statistical iterative reconstruction. Br. J. Radiol. 2011;84(1000):335–341. 21081572 10.1259/bjr/57998586 Kalra MK Homayounieh F Arru C : Chest CT practice and protocols for COVID-19 from radiation dose management perspective. Eur. Radiol. 2020;30(12):6554–6560. 32621238 10.1007/s00330-020-07034-x Karabulut N Törü M Gelebek V : Comparison of low-dose and standard-dose helical CT in the evaluation of pulmonary nodules. Eur. Radiol. 2002;12(11):2764–2769. 12386771 10.1007/s00330-002-1368-4 Kashyape R Jain R : The utility of HRCT in the initial diagnosis of COVID-19 pneumonia-An Indian perspective. Indian J. Radiol. Imaging. 2021;31(5):S178–S181. 33814779 10.4103/ijri. IJRI_944_20 PMC7996707 Katsura M Matsuda I Akahane M : Model-based iterative reconstruction technique for radiation dose reduction in chest CT: Comparison with the adaptive statistical iterative reconstruction technique. Eur. Radiol. 2012;22(8):1613–1623. 22538629 10.1007/s00330-012-2452-z Kumar M Pai KM Vineetha R : Occupation-related musculoskeletal disorders among dental professionals. Med. Pharm. Rep. 2020;93(4):405–409. 33225267 10.15386/mpr-1581 Li Y Xia L : Coronavirus disease 2019 (COVID-19): Role of chest CT in diagnosis and management. Am. J. Roentgenol. 2020;214(6):1280–1286. 10.2214/AJR.20.22954 Mahdavi A Haseli S Mahdavi A : The role of repeat chest CT scan in the COVID-19 Pandemic. Acad. Radiol. 2020;27(7):1049–1050. 32389601 10.1016/j.acra.2020.04.031 Menzel HG Schibilla H Teunen D : European guidelines on quality criteria for computed tomography. Luxembourg: European Commission;2000;16262. Mullenders L Atkinson M Paretzke H : Assessing cancer risks of low-dose radiation. Nat. Rev. Cancer. 2009;9(8):596–604. 10.1038/nrc2677 Neroladaki A Botsikas D Boudabbous S : Computed tomography of the chest with model-based iterative reconstruction using a radiation exposure similar to chest X-ray examination: Preliminary observations. Eur. Radiol. 2013;23(2):360–366. 22892722 10.1007/s00330-012-2627-7 Rubin GD Ryerson CJ Haramati LB : The Role of Chest Imaging in Patient Management During the COVID-19 Pandemic: A Multinational Consensus Statement From the Fleischner Society. Chest. 2020;158(1):106–116. 32275978 10.1016/j.chest.2020.04.003 Salvi S Kumar GA Dhaliwal RS : The burden of chronic respiratory diseases and their heterogeneity across the states of India: the Global Burden of Disease Study 1990–2016. Lancet Glob. Health. 2018;6(12):e1363–e1374. 30219316 10.1016/S2214-109X(18)30409-1 Shah DJ Sachs RK Wilson DJ : Radiation-induced cancer: A modern view. Br. J. Radiol. 2012;85(1020):e1166–e1173. 23175483 10.1259/bjr/25026140 Sukumar S : Low dose protocol for high resolution CT thorax: influence of matrix size and tube voltage on image quality and radiation dose. 2022. Harvard Dataverse, V1. 10.7910/DVN/P3BL0F To T Viegi G Cruz A : A global respiratory perspective on the COVID-19 pandemic: Commentary and action proposals. Eur. Respir. J. 2020;56(1):2001704. 32586874 10.1183/13993003.01704-2020 Tsukada J Yamada M Yamada Y : Comparison of the diagnostic accuracy of FBP, ASiR, and MBIR reconstruction during CT angiography in the evaluation of a vessel phantom with calcified stenosis in a distal superficial femoral artery in a cadaver extremity. Medicine (United States). 2016;95(27):e4127. 27399123 10.1097/MD.0000000000004127 Wang YXJ Liu W-H Yang M : The role of CT for Covid-19 patient’s management remains poorly defined. Ann. Transl. Med. 2020;8(4):145–145. 32175437 10.21037/atm.2020.02.71 Yanagawa M Gyobu T Leung AN : Ultra-low-dose CT of the Lung. Effect of Iterative Reconstruction Techniques on Image Quality. Acad. Radiol. 2014;21(6):695–703. 10.1016/j.acra.2014.01.023 Yeung AWK : The “as Low As Reasonably Achievable” (ALARA) principle: A brief historical overview and a bibliometric analysis of the most cited publications. Radioprotection. 2019;54(2):103–109. 10.1051/radiopro/2019016 10.5256/f1000research.122637.r181026 Reviewer response for version 1 Mei Kai 1 Referee https://orcid.org/0000-0002-6429-7830 University of Pennsylvania, Philadelphia, Pennsylvania, USA

Competing interests: No competing interests were disclosed.

 13 9 2024 Copyright: © 2024 Mei K 2024 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation reject

This paper is relatively well written, and the study was adequately well performed. The main problem of this paper is the motivation of design.

The only parameters in a CT protocol that affect radiation dose are tube voltage, tube current, rotation speed and helical pitch. Matrix size has no effect on radiation dose and has minimal effect on diagnosis quality, compared to reconstruction filter: radiologist can always opt to sharper filters.

The tuning of tube voltage is mostly determined by the nature of the examination and the property of the X-ray source of that CT scanner, it is not used as a mean to reduce radiation dose. Tube voltage changes the contrast of the CT image, so specific examinations need to be done in specific tube voltage.

In addition, the matrix size is not really relevant in dose studies. It only changes how you store the patient data. There are so much more other ways to enhance the diagnosis quality (from reconstruction filters to post extrapolations in the DICOM viewer), when the matrix size is limited.

The current topic of this manuscript lacks interest in CT research fields and I would recommend the author to redesign the study to compare with pitch, scanning speed and tube current etc. instead.

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

Partly

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

CT researcher for over 8 years. CT reconstruction, multi-energy CT, reconstruction algorithms, clinical CT protocols, CT dose reduction, low-dose CT algorithms.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.

 10.5256/f1000research.122637.r181038 Reviewer response for version 1 Rodesch Pierre-Antoine 1 Referee https://orcid.org/0000-0001-6199-0042 University of Victoria, Victoria, British Columbia, Canada

Competing interests: No competing interests were disclosed.

 7 7 2023 Copyright: © 2023 Rodesch PA 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

The authors are studying the optimization of the dose in the frame of high-resolution computed tomography (HRCT) thorax imaging. Given the advances in iterative reconstruction (IR), new possibilities to lower the dose and increase the diagnostic score can be investigated. Indeed, IR can compensate for the noise increase at lower doses. One possibility to lower the dose while maintaining the same image quality is to use a different source voltage. Here, the authors investigate two different voltages (100 and 120 kVp). In order to increase the diagnostic score, the authors study three different matrix sizes (512, 718 and 1024) to increase the spatial resolution. The combination of different voltages and matrix size are also studied to provide 6 different protocols. The authors use phantom data and a large cohort of patients to compare the 6 protocols. This represents enough data to make the work relevant and make recommendations.

The work is rigorously written and relevant to the field. The context and objective are nicely described and very actual with appropriate citations.

However, I would have designed the methodology differently, especially on the phantom acquisition. It would be more interesting to be able to have cases with the same dose, or the same noise, or the same CNR, which is not the case for any combination of protocols. Here the authors present an increased CNR and diagnostic score for a higher dose, which is predictable and can not be used to draw conclusions.

As the matrix size increases, the noise increases as well. The IDose level should be increased for the 718 and 1024 cases to match the noise level.

As the voltage decreases, the contrast is expected to increase, and it would be possible to increase the IDose level as well to match the CNR. (The other possibility would be to increase the current to match the dose).

 In table 2, instead of the SNR, the noise and contrast value would be more relevant to highlight the noise and contrast differences.

In table 3, it would be easier to present the results in a different way: by identifying the lowest value and calculate all percentages in comparison to the latter.

Table 4 and Figure 3 are redundant, I think you could only keep Figure 3 and would include the scores (Table 6) in this figure. Why is there no p-values computed for the scores?

Form Table 6, we conclude that Protocol B is the best, but you recommend protocol D, it is not clear why: why not recommending protocol E? I understand that redoing the patient acquisitions/reconstructions is not feasible. But to support your affirmation, you will need to provide more results. Could you calculate a normalized diagnostic score by dividing it by the dose, in a similar way as the FOM? Could you also provide the percentage of cases for each protocol that presented a score lower than 4 (or 4.5?) and so would not be acceptable? Please elaborate the discussion.

Minor comments:

I am not sure using A, B, C, D, E and F is the best way to refer to each protocol, you should use 120/512, 120/718, etc… It does not take more space and help to facilitate the comprehension.

In the abstract you don’t mention the FOV, which is important when the matrix size varies.

In this abstract sentence “For quantitative image quality, the difference noted was also statistically significant among two kVp settings.”, I don’t understand which kVp performs the best.

In material and methods, you don’t indicate the exposure (mAs) parameter in either phantom or patient acquisition.

In material and methods, you don’t explain the different patient groups: I was expecting to reconstruct each patient with 3 different matrix size but that was not the case, right? (Only one matrix size per patient?)

You could also mention photon-counting CT that could reduce further the dose of innovative denoising techniques such as deep learning methods.

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Yes

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

No

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

medical CT imaging, image quality metrics, thorax imaging, photon-counting CT.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

 10.5256/f1000research.122637.r130052 Reviewer response for version 1 Kotian Rahul P. 1 Referee Medical Imaging Sciences, Gulf Medical University, Ajman, United Arab Emirates

Competing interests: No competing interests were disclosed.

 28 4 2022 Copyright: © 2022 Kotian RP 2022 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve

This paper reports India's National CT radiation dose and image quality related to HRCT Thorax scans. I think this is an important work that might motivate other "smaller" countries to do similar initiatives. Therefore I find it relevant and good to index(even if it is not the newest thing in the world scenario related to CT radiation dose). The paper is well written, the introduction section is crisp and addresses the current scenario of radiation dose and image quality in HRCT with different matrix combinations. The materials and methods section clearly defines the steps taken in data collection using a 32cm body phantom. Lastly, the results obtained are valuable and should be utilized as a routine imaging protocol to obtain optimum image quality with significant dose reduction in HRCT Thorax scans.

It would add value and improve the visibility of the current manuscript if the authors decide to cite the following high impact articles published in the area of CT Dose reduction and image quality:

Rawashdeh M, McEntee M, Zaitoun M,  et al. Knowledge and practice of computed tomography exposure parameters amongst radiographers in Jordan. 1

Mohamad F, Lina K, Ghida A, et al. National diagnostic reference levels have a lot of potential but a long way to go. A systematic review on the current status of adult diagnostic reference levels in head, chest and abdominopelvic Computed Tomography 2

Rawashdeh M, Abdelrahman M, Zaitoun M, et al. Diagnostic reference levels for paediatric CT in Jordan. 3

Rawashdeh M, Saade C, Zaitoun M, et al. Establishment of diagnostic reference levels in cardiac computed tomography. 4

Rawashdeh, M, Saade C. Radiation dose reduction considerations and imaging patterns of ground glass opacities in coronavirus: risk of over exposure in computed tomography.  4

El-Merhi F, Mohamad M, Haydar A, et al. Qualitative and quantitative radiological analysis of non-contrast CT is a strong indicator in patients with acute pyelonephritis. 5

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

Yes

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Yes

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

MRI, Fractional Anisotropy, Diffusion tensor imaging, Parkinson’s Disease, Computed Tomography Radiation Dose, Computed Tomography

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

References : Knowledge and practice of computed tomography exposure parameters amongst radiographers in Jordan. Computers in Biology and Medicine .2018;102: 10.1016/j.compbiomed.2018.09.020 132-137 10.1016/j.compbiomed.2018.09.020 : A systematic review on the current status of adult diagnostic reference levels in head, chest and abdominopelvic Computed Tomography. J Radiol Prot .2020;40(3) : 10.1088/1361-6498/ab826f R71-R98 32203948 10.1088/1361-6498/ab826f : Diagnostic reference levels for paediatric CT in Jordan. J Radiol Prot .2019;39(4) : 10.1088/1361-6498/ab3ee2 1060-1073 31469115 10.1088/1361-6498/ab3ee2 : Establishment of diagnostic reference levels in cardiac computed tomography. J Appl Clin Med Phys .2019;20(10) : 10.1002/acm2.12711 181-186 31469229 10.1002/acm2.12711 : Radiation dose reduction considerations and imaging patterns of ground glass opacities in coronavirus: risk of over exposure in computed tomography. Radiol Med .2021;126(3) : 10.1007/s11547-020-01271-2 380-387 32897493 10.1007/s11547-020-01271-2 : Qualitative and quantitative radiological analysis of non-contrast CT is a strong indicator in patients with acute pyelonephritis. Am J Emerg Med .2018;36(4) : 10.1016/j.ajem.2017.09.026 589-593 29055617 10.1016/j.ajem.2017.09.026