Comparison of image quality between Deep learning image reconstruction and Iterative reconstruction technique for CT Brain- a pilot study

Background Non-contrast Computed Tomography (NCCT) plays a pivotal role in assessing central nervous system disorders and is a crucial diagnostic method. Iterative reconstruction (IR) methods have enhanced image quality (IQ) but may result in a blotchy appearance and decreased resolution for subtle contrasts. The deep-learning image reconstruction (DLIR) algorithm, which integrates a convolutional neural network (CNN) into the reconstruction process, generates high-quality images with minimal noise. Hence, the objective of this study was to assess the IQ of the Precise Image (DLIR) and the IR technique (iDose 4) for the NCCT brain. Methods This is a prospective study. Thirty patients who underwent NCCT brain were included. The images were reconstructed using DLIR-standard and iDose 4. Qualitative IQ analysis parameters, such as overall image quality (OQ), subjective image noise (SIN), and artifacts, were measured. Quantitative IQ analysis parameters such as Computed Tomography (CT) attenuation (HU), image noise (IN), posterior fossa index (PFI), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) in the basal ganglia (BG) and centrum-semiovale (CSO) were measured. Paired t-tests were performed for qualitative and quantitative IQ analyses between the iDose 4 and DLIR-standard. Kappa statistics were used to assess inter-observer agreement for qualitative analysis. Results Quantitative IQ analysis showed significant differences (p<0.05) in IN, SNR, and CNR between the iDose 4 and DLIR-standard at the BG and CSO levels. IN was reduced (41.8-47.6%), SNR (65-82%), and CNR (68-78.8%) were increased with DLIR-standard. PFI was reduced (27.08%) the DLIR-standard. Qualitative IQ analysis showed significant differences (p<0.05) in OQ, SIN, and artifacts between the DLIR standard and iDose 4. The DLIR standard showed higher qualitative IQ scores than the iDose 4. Conclusion DLIR standard yielded superior quantitative and qualitative IQ compared to the IR technique (iDose4). The DLIR-standard significantly reduced the IN and artifacts compared to iDose 4 in the NCCT brain.


Introduction
Computed tomography (CT) is the primary imaging modality used to evaluate patients suspected to have central nervous system disorders.The ability to visualize brain regions quickly and thoroughly is one of their main advantages.This helps in the timely assessment of problems, including stroke, trauma, and intracranial lesions.[3] One notable development in CT reconstruction technology is the use of iterative reconstruction (IR) techniques.iDose 4 is a 4 th generation IR method released by Philips Healthcare that offers improved image quality (IQ) with a reduced radiation dose.IQ and diagnostic precision are improved through IR, which uses sophisticated mathematical techniques to optimize and refine image data.A significant reduction in image noise (IN) allows for clearer visualization of anatomical structures, especially in regions of low contrast, which is the main advantage of the IR technique.IR allows the acquisition of high-quality images with a decreased radiation dose (RD) for patients.For vulnerable groups such as children or those who are radiation-sensitive, this is especially important. 4,5[8] Deep learning image reconstruction (DLIR) algorithms represent a transformative leap in image reconstruction and dose reduction in CT.DLIR in Philips Healthcare is called a Precise Image (PI).Philips PI is the latest and most reliable way to reconstruct high-quality CT images using artificial intelligence (AI) techniques.A trained deep learning neural network was used in the PI reconstruction process.With the fastest reconstruction speed in the market, PI preserves the traditional view of FBP photos. 9By harnessing the capabilities of artificial intelligence (AI), particularly convolutional neural networks (CNNs), these algorithms revolutionize image reconstruction by learning intricate patterns from raw CT data.These algorithms learn complex relationships between sparse or noisy input projections and corresponding standard-dose reference images, enabling the generation of clinically acceptable reconstructions even when using a reduced radiation dose.By leveraging the inherent information within the data and learning intricate patterns, DLIR contributes to the advancement of dose reduction strategies in CT.1][12] There are limited studies on the usefulness of DLIR on IQ in the NCCT brain.Hence, the aim of this study was to compare the IQ between the new Precise Image (DLIR) and IR (iDose 4 ) techniques for the NCCT Brain.

Methods
This is a prospective study.The Institutional Ethical Committee (IEC 400/2022) was obtained from Kasturba Medical College and Hospital, Manipal, India, on 1 st July 1, 2023, followed by the Clinical Trial Registry -India (CTRI) registration (CTRI/2023/07/055310) on 18 th July 2023.Written informed consent was obtained from all the participants for publication and participation in the study.
Eligibility criteria: Thirty patients referred for the NCCT brain were included.Patients referred to the NCCT brain for various clinical indications such as trauma, seizures, stroke, headache, vomiting, fever, chills, and breast carcinoma were included.Patients who were uncooperative and who underwent CT scans with motion artifacts were excluded.The patients included in the study had neuropathological findings on CT, including hemorrhage (n=10), infarct (n=10), tumor (n = 5), VP shunt (n = 1), arachnoid cyst (n = 1), metastases with edema (n=1), cerebral atrophy (n=1), and encephalomalacia (n=1).
CT Image acquisition: This study was performed at the Department of Radiodiagnosis, Kasturba Medical College and Hospital.Patients referred for NCCT brain examinations underwent 128 slice CT (Incisive, Philips Healthcare).The technical parameters for the NCCT brain acquisition are listed in Table 1.The images were reconstructed using iDose 4 (level 3) 4 and the DLIR reconstruction 9 level standard.
"Qualitative Image quality analysis": Two radiologists [reader 1 (R1) and reader 2 (R2)] with over 15 years of experience in neuroradiology imaging evaluated the CT images.Both the readers were blinded to the reconstruction level.The readers assessed the "Overall image quality" (OQ), "Image noise" (IN), "Artifacts" using 5-point Likert scale (Figure 1)."Quantitative Image quality analysis": CT attenuation (HU) and image noise (IN) of gray matter (GM) and white matter (WM) at the level of the basal ganglia (BG) and centrum-semiovale (CSO) regions were measured.To calculate attenuation (HU) at the GM and WM of the BG, an ROI of 0.1-0.2cm 2 was placed in the thalamus and posterior limb of the internal capsule (PIC).For calculating attenuation at GM and WM of CSO, the ROI of 0.1-0.2cm 2 was placed in the region of frontal WM and adjacent cortical GM (Figure 2a-b).
The posterior fossa index (PFI) was calculated as the image noise (Standard deviation-SD) of the HU values in the pons.To calculate the PFI, an ROI of 0.2-0.3cm 2 was placed in the pons region of the posterior cranial fossa (Figure 2c).
The signal-to-noise ratio (SNR) at the BG and CSO levels was calculated as mean CT attenuation (HU)/standard deviation (SD) (SD: Image noise).
The contrast-to-noise ratio (CNR) at the BG and CSO levels was calculated using the following formula:

CNR ¼
Radiation dose metrics such as "CTDI volume (CTDI vol ), dose length product (DLP), and size-specific dose estimate (SSDE)" were recorded from the display of the console monitor.
Statistical analysis SPSS (IBM, V20.0) was used for statistical analysis.Paired t-tests were performed for qualitative and quantitative IQ analyses between the iDose 4 and DLIR-standard."Kappa (k) statistics" were used to check the interobserver agreement for qualitative analysis.The k-value was considered as follows: <0.20, poor agreement, 0.21-0.40-Fair agreement, 0.41-0.60-Moderate agreement, 0.61-0.80-Good agreement, 0.81-1.00-Excellent agreement".Statistical significance was set at P < 0.05.

Qualitative IQ analysis
Qualitative IQ analysis showed an increase in scores for OQ, IN, and artifacts with DLIR-standard compared to iDose 4 for both readers (Table 3) (Figure 3a-b).

Interobserver agreement
For OQ, the agreement between the readers was excellent for iDose 4 (0.902) and the DLIR-standard (0.821).For IN, the agreement between readers was excellent for iDose 4 (1.00) and the DLIR-Standard (0.837).For artifacts, the agreement between the readers was good for iDose 4 (0.80) and excellent for the DLIR-Standard (0.943).
Quantitative IQ analysis CT Attenuation (HU) at the BG and CSO levels did not show significant differences (<0.05) in the GM thalamus, WM PIC, adjacent cortical GM, and frontal WM between the iDose 4 and DLIR-standard (Table 4).IN showed significant differences (<0.05) between iDose 4 and DLIR-standard at the BG level (GM thalamus, WM PIC) and CSO level   (adjacent cortical GM, frontal WM).IN showed 42.8% and 43.47% decreases in GM thalamus and WM PIC, respectively, with DLIR-standard compared to iDose 4 .IN showed 41.86% and 47.61% decrease in adjacent cortical GM and frontal WM, respectively, with DLIR-standard compared to iDose 4 .PFI showed significant difference (<0.05) between iDose 4 and DLIR-standard with 27.08% IN reduction in the pons region with DLIR-standard compared to iDose 4 .SNR at BG and CSO levels showed significant differences (<0.05) for the GM thalamus, WM PIC, adjacent cortical GM, and frontal WM between the iDose 4 and DLIR-standard.SNR showed 67.60% and 76.78% increases in GM thalamus and WM PIC, respectively, with DLIR-standard compared to iDose 4 .SNR showed 65% and 82.81% increases at adjacent cortical GM and frontal WM, respectively, with DLIR-standard compared to iDose 4 .CNR at BG and CSO levels showed significant differences (p < 0.05) in GM thalamus and WM PIC differentiation, adjacent cortical GM, and frontal WM differentiation between iDose 4 and DLIR-standard.CNR showed 68% and 78.8% increases in BG and CSO, respectively, with DLIR-standard compared to iDose 4 .

Discussion
In the present study, we compared the qualitative and quantitative IQ between the DLIR-standard (Precise Image) and IR (iDose 4 ) techniques for the NCCT brain.Our study noticed that both qualitative and quantitative IQ improved significantly with the DLIR-standard compared with the iDose 4 .The new DLIR technique, Precise Image, outperformed the IR technique (iDose 4 ).Our study found that the DLIR standard showed a significant reduction in IN and an increase in SNR and CNR at BG and CSO levels.The DLIR-standard showed higher subjective IQ scores with excellent agreement between readers compared to the iDose 4 .The lower IN, higher SNR, and CNR might allow for lowering the radiation dose with the DLIR-standard compared to iDose 4 for the NCCT brain.
Studies by Kim et al. 13 and Alagic et al. 14 showed 24-52% and 3.5-43% reduction in IN for NCCT brains with DLIR (True Fidelity; GE) reconstruction levels of low, medium, and high compared with ASIR-V (Adaptive statistical iterative reconstruction-Veo) at BG and CSO levels.DLIR-standard in the present showed a 41.8-47.6%reduction in IN at BG and CSO levels, which is similar to the results of Kim et al. 13 and Alagic et al. 14 Another Two studies by Oostveen et al. 15  For the NCCT brain, diagnostic evaluation of the posterior fossa in emergency situations to identify hemorrhagic and ischemic events is important.However, the posterior cranial fossa often experiences beam hardening, streak, and partial volume artifacts due to the presence of bony structures surrounding the cerebellum, pons, and medulla, which leads to diagnostic challenges in identifying hemorrhage and infarct in this region.The artifact index could indicate the extent of CT number fluctuations resulting from artifacts, along with intrinsic image noise linked to factors related to both the scanner and the patient. 17,18Our study noticed a 27.08% reduction in IN and artifact index in the pons region of the posterior fossa with DLIR-standard compared to iDose 4 which was similar to the artifact index reported by Kim et al. 13 (17-38%) and Alagic et al. 14 (6.8-32.8%).However, Cozzi et al. 16 reported a higher artifact index (median 8.4; interquartile range 7.3-9.2) with DLIR (AiCE) than with AIDR-3D (median 7.5; interquartile range 6.9-8.3) in thin sections, which is contrary to the study reported by Oostveen et al. 15 with the same DLIR technique.The reason for this could be the difference in the placement of the ROI and slice thickness used between the two studies.
Our study observed higher SNR (65-82%) at BG and CSO levels with DLIR-standard, which suggests better gray and white matter differentiation compared to iDose 4 .The findings of our study were similar to the results of Alagic et al. 14 (2-89%) with DLIR-low, medium, and high levels, and Pula et al. 19 (46-59%) with DLIR-High compared to IR techniques.A study by Oostveen et al. 15 reported a slightly lower reduction in SNR (5-26%) compared to our study because of the difference in the formula used for calculating the SNR.Our study observed an increase in CNR (68-78.8%)with DLIR-standard at BG and CSO levels, which suggested better gray and white matter differentiation compared to iDose 4 .The results of our study were similar to the findings reported by Alagic et al. 14 (2.4-53%) and Cozzi et al. 16 (28-39%).However, Kim et al. 13 reported an increase in CNR of 99% with DLIR-high DLIR.
Our study found no significant difference (p>0.05) in CT attenuation (HU) between DLIR-standard and iDose 4 at the BG and CSO levels.The findings of our study are similar to the results of Kim et al. 13 which showed no significant difference in CT attenuation of GM between DLIR levels and IR technique.However, a study by Alagic et al. 14 reported significant differences in CT attenuation between DLIR levels and IR technique at the PLIC WM and adjacent cortical GM, which might suggest that DLIR could lead to minor changes in attenuation values; however, this finding is unlikely to have significant clinical consequences.
The DLIR-standard showed higher qualitative scores for OQ, IN, and artifacts compared to iDose 4 which is similar to the findings reported by Oostveen et al. 15 and Pula et al. 19 Studies by Kim et al. 13 and Alagic et al. 14 reported an increase in qualitative scores with an increase in the strengths of DLIR from low to high compared to IR techniques.
Our study has a few limitations.First, the study involved a small sample size, and it is necessary to conduct further research with a larger patient cohort to confirm our study findings.Second, the study did not directly evaluate the diagnostic efficacy, which is a crucial step in understanding the complete clinical advantages of DLIR.Third, CT scanning was performed using a standard dose protocol, making it challenging to directly ascertain the potential dosereduction benefits of DLIR.

Conclusion
The New DLIR Precise Image (DLIR) technique offers improved image quality with reduced image noise and higher SNR and CNR than iDose 4 .The DLIR standard also showed higher qualitative image quality scores than the iDose 4 .The reduction of posterior fossa artifacts with the DLIR standard for the NCCT brain improves the diagnostic accuracy of identifying hemorrhages/infarcts in emergency cases.Our current study may provide implications for performing lowdose scans with reduced radiation doses using DLIR in the NCCT brain.

Mustapha Barde
Bayero University, Kano, Kano, Nigeria The article revealed the effectiveness of deep learning iterative image reconstruction (DLIR) as compared to iDose4 (iterative reconstruction) in terms of qualitative and quantitative image quality (IQ) in non-contrast brain CT studies.DLIR presented better signal-to-noise ratio, contrast to noise ratio and reduced image noise.it enhances radiation dose optimization by using low radiation dose for purpose of image acquisition.

The article is worthy of being published upon addressing the below minor corrections Minor corrections
The abbreviation FBP used in the introduction and VP in the eligibility criteria part should be written in full for the first time, and subsequently they can be used.

1.
In figure 2. The numbers 1 and 2 (ROIs) should be defined on either the image or the figure description.

2.
Formula should be written in a standard format, where superscripts is required, they should be correctly inserted.

3.
Based on the qualitative IQ analysis results of the different observers, the second observer R2 results should be explained with the corresponding P-value as well, as shown in Table 3.
Try to modify the table to indicate the P-values based on each observer.

4.
Quantitative image analysis: the CT attenuation (HU) reported p-value in the result, did not correspond to Table 4. check for Possible sign error in the p-value stated.

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

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Image analysis and processing, CT Dosimetry 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.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Yes

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

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Medical imaging, radiography, radiation protection 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.

Major comments
This research article provides a comprehensive comparison of image quality parameters, such as diagnostic value, image noise and artifacts. of DLIR over traditional iterative reconstruction methods in Non contrast CT brain examinations.This will benefit the technologist and the radiologist to understand the various benefits using various available reconstruction method in Computed Tomography particularly in the Non contrast CT brain.Since this research was done in a mixed method (both qualitative and quantitative), it reveals a valuable result to upgrade the reconstruction algorithm in future.

Jerald Paul
Fatima College of Health Sciences, Al Ain, United Arab Emirates

Major Comments:
This is an interesting study dealing with CT image quality using DLIR and iDose 4 for non-contrast CT brain.The study concludes that DLIR showed higher qualitative scores and reduced image noise (41.8-47.6%)with higher SNR (65-82%) and CNR (68-78.8%).
The innovation of the study is good.The authors clearly describe the scientific rationale or hypothesis of this study in the manuscript.Methods section is described well with various image quality measurements and its formula.Results explained clearly.Discussion compared the image quality measures from present study with references from literature.Conclusion explained the benefits of DLIR in Non-contrast CT brain.

Minor comments:
The Methods section can include more details about DLIR.
Reference for using CNR formula can be provided.Abbreviations like SPSS in statistical section can be provided.

Key Findings and Analysis
The study provides a comprehensive comparison of image quality parameters, both qualitative and quantitative, which are crucial for radiographers, technologists, and radiologists seeking to understand the tangible benefits of DLIR over traditional iterative reconstruction methods in CT brain examinations.
A particularly compelling aspect highlighted in the article is the measurement of the posterior fossa artifact index.Historically, posterior fossa beam hardening artifacts have posed challenges in CT brain examinations, particularly in assessing conditions such as posterior fossa contusions and infarcts.The reduction in artifacts observed in DLIR images compared to previous methods like iDose4 represents a significant advancement that aids radiologists in making more accurate diagnoses.Moreover, the article suggests that DLIR has the potential to lower radiation doses while maintaining diagnostic quality, which is critical for the safety and effectiveness of follow-up head CT scans.

Minor Comments and Future Directions
While the study is commendable in its depth and scope, minor adjustments such as limiting the use of quotation marks in the manuscript could enhance clarity and readability.Future research avenues could explore the application of DLIR in other body parts and in low-dose CT examinations to further elucidate its benefits and effectiveness across different clinical scenarios.
This research article contributes valuable insights into the applications and benefits of AI-based technologies in CT for improving image quality and optimizing radiation dose.The adoption of AI techniques such as DLIR, as reported in this study, represents a significant advancement in enhancing patient care through more precise diagnostic imaging.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Yes

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate?I cannot comment.A qualified statistician is required.
Are all the source data underlying the results available to ensure full reproducibility?Yes

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.

Reviewer Expertise: Radiological Physics
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.
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com

Figure 1 .
Figure 1.Showing the 5-point Likert scale for qualitative image quality analysis.

Figure 2 .
Figure 2. Axial CT images (DLIR-standard) of 61-year old male at the level of centrum semiovale (a) and basal ganglia (b) showing region of interest (ROI) in gray matter and white matter.Axial CT image at the posterior cranial fossa (c) with ROI drawn in the pons region between the petrous bone.

Figure 3 .
Figure 3. Axial CT images showing the improved image quality with (a) iDose 4 compared to (b) DLIR-standard.

Reviewer
Report 03 July 2024 https://doi.org/10.5256/f1000research.165371.r296706© 2024 Shetty S. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Shashi Kumar Shetty NITTE university, Karnataka, India Please find the comments for the article Major comments Introduction: Well-structured.It explained the current reconstruction techniques of CT available in the market and the disadvantages of iDose 4 .Deep learning-based reconstruction techniques are currently being validated for various applications in CT for image quality improvement and radiation dose reduction.Hence the aim and research gap addressed here is very much relevant to the current context of issues to be addressed while using DLIR techniques for clinical use.Methodology: It clearly explains the key concepts such as inclusion, exclusion criteria, qualitative and quantitative image analysis.Statistical analysis: The tests used for comparing the image quality measures and interobserver agreement appear appropriate.Results: Well-explained with descriptive statistics as well as p-values.The tables provided are informative and easy to understand.Discussed: Clearly explained the results and implications of DLIR in image noise and artifact reduction with clinical relevance.Conclusion: It summarizes the key findings and applications of new precise image technique in clinical aspect for CT brain Minor comments Are there any additional costs required for installing these DLIR techniques in CT machines?Abbreviations needs to be provided at its first use Overall, the original research paper provides a significant contribution to the field of advancements in image reconstruction techniques in CT.

Reviewer Report 03
July 2024 https://doi.org/10.5256/f1000research.165371.r296708© 2024 S D. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Dr. Tamijeselvan S College of MRIT, Mother Teresa Postgraduate and Research institute of Health sciences, Indira Nagar, Puducherry, India In this article the author(s) compared the image quality obtained by 2 reconstruction methods namely DLIR standard and IR technique (iDose4).The findings revealed that DLIR-standard gives the better quality image by reducing the image noise and artifacts compared to iDose4 in the NCCT brain.

©
2024 Palaniyappan S. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Senthil Manikandan Palaniyappan Assistant professor, Department of Radiation Physics, Kidwai Memorial Institute of Oncology, Bangalore, India In recent years, significant advancements have been made in CT image reconstruction technology, particularly with the introduction of deep learning-based image reconstruction (DLIR) algorithms.One notable development is the "Precise Image" by Philips Healthcare, which aims to reduce image noise and enhance overall image quality for Non-contrast CT brain scans.

Table 1 .
Showing the CT technical parameters for Non contrast CT Brain.

Table 2 .
Showing the characteristics of population.

Table 3 .
Showing qualitative image quality analysis between iDose 4 and DLIR-standard.
OQ -Over all image quality, IN -Image noise, DLIR -Deep learning image reconstruction.

Table 4 .
Comparison of quantitative image quality analysis between iDose 4 and DLIR-standard.
GM -gray matter; HU -Hounsfield unit; WM -White matter; PIC -Posterior limb of the internal capsule.

Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Yes Are sufficient details of methods and analysis provided to allow replication by others? 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 Are the conclusions drawn adequately supported by the results? Yes Competing Interests:
The biggest challenge in avoiding the base of skull artifacts in CT brain examinations is taken into account for the comparison.Any technical advancement which reduces the patient dose and improve the diagnostic value is always useful in diagnostic radiology.At this view this comparison study is very much appreciable and need of the hour.Minor commentsMethod of sample selection (how the researcher select the sample from the population) can be included in the Methods section.Instead of 2 radiologist more number can be used to get more appropriate results.Otherwise as a whole, this research article provides valuable data which will lead to quality imaging with less patient dose References 1. Alagic Z, Diaz Cardenas J, Halldorsson K, Grozman V, et al.: Deep learning versus iterative image reconstruction algorithm for head CT in trauma.Emerg Radiol.2022; 29 (2): 339-352 PubMed Abstract | Publisher Full Text No competing interests were disclosed.

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Yes Are sufficient details of methods and analysis provided to allow replication by others? 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 Are the conclusions drawn adequately supported by the results? Yes Competing Interests:
The article is very attractive and provides deep insights about deep learning-based reconstruction techniques in CT.No competing interests were disclosed. ÂIs

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
https://doi.org/10.5256/f1000research.165371.r296709