Open AccessArticle Deep Learning Reconstruction Specialized for Inner Ear: Improving Image Quality and Anatomical Structure Visualization as Compared with Conventional Hybrid-Type Iterative Reconstruction on High-Definition CT 1 Department of Diagnostic Radiology, School of Medicine, Fujita Health University, 1-98 Dengakugakubo, Kutsukake-cho, Toyoake 470-1192, Aichi, Japan 2 Joint Research Laboratory of Advanced Medical Imaging and Artificial Intelligence, School of Medicine, Fujita Health University, Toyoake 470-1192, Aichi, Japan 3 Canon Medical Systems Corporation, Otawara 324-8550, Tochigi, Japan 4 Department of Radiology, Fujita Health University Bantane Hospital, Nagoya 454-8509, Aichi, Japan 5 Department of Radiology, Fujita Health University Okazaki Medical Center, Okazaki 444-0827, Aichi, Japan * Author to whom correspondence should be addressed. Diagnostics 2026, 16(12), 1756; https://doi.org/10.3390/diagnostics16121756 (registering DOI) Submission received: 5 May 2026 / Revised: 1 June 2026 / Accepted: 5 June 2026 / Published: 6 June 2026 Background/Objectives: To directly compare the capabilities of hybrid-type iterative reconstruction (IR) with the newly developed deep learning reconstruction (DLR) for the inner ear on high-definition CT (HDCT) obtained using the super-high-resolution (SHR) mode for external, middle and inner ear evaluations and diagnosis in patients with and without otologic diseases. Methods: Included in this study were 140 patients who had undergone HDCT, consisting of 32 otologic disease patients and 108 non-otologic disease patients, and 280 inner and middle ears and temporal bones were evaluated on a per ear analysis. Signal-to-noise ratios (SNRs) of the temporal bone surrounding the aural vestibule of the ear and in the vestibule as well as the cerebellar hemisphere, overall image and detailed evaluation of the visibility of anatomical landmarks in the middle and inner ear and temporal bone obtained with the two methods were assessed and statistically compared using the paired t-test or Wilcoxon’s signed-rank test. Then, receiver operating characteristic (ROC) analysis was performed to compare diagnostic performance between two reconstruction methods. Results: Each SNR of DLR was significantly higher than that of hybrid-type IR ( p 0.05). 4. Discussion Our study results indicate that newly developed DLR, when compared with clinically applied hybrid-type IR on HDCT, could improve image quality and detailed evaluations of visibility of anatomical landmarks in middle and inner ears and temporal bone, although diagnostic performance was not affected. To the best of our knowledge, this study was the first to compare the capability of DLR for image quality and detailed evaluations of visibility of anatomical landmarks in middle and inner ears and temporal bone with that of hybrid-type IR on HDCT. All κ values for each of the inter-observer agreements ranged between 0.61 and 1.0, and all inter-observer agreements were determined as ‘substantial’ or ‘almost perfect’. Our evaluations in this study can therefore be considered reproducible [ 23]. A comparison of image quality, artifacts and detailed evaluation of the visibility of anatomical landmarks in middle and inner ears and the temporal bone showed the use of DLR could result in significant improvement as compared with that of hybrid-type IR on HDCT. However, when comparing diagnostic performance between the above-mentioned two methods, there were no significant differences in AUC, sensitivity, specificity and accuracy between them. These results suggested that improved quantitative image quality gains with little clinical benefit for diagnosis were due to the following reasons: hybrid-type IR has already provided good image quality for clinical diagnosis, and DLR can reduce image noise more than hybrid-type IR and results in providing a sharper or clearer image as compared with DLR. Therefore, DLR may improve the confidence level for diagnosis level more than hybrid-type IR, although there was no significant difference in AUC between the two methods. Therefore, these facts suggest that DLR merits the use of HDCT data for improving image quality and visualization of anatomical landmarks in middle and inner ears and temporal bones, although the newly developed method had less influence on the diagnosis of otologic diseases in routine clinical practice. However, improved SNR on HDCT reconstructed with DLR may be possible to reduce radiation dose more than that with hybrid-type IR while keeping image quality and diagnostic performance. Therefore, DLR may impact radiation dose reduction as compared with hybrid-type IR, and further investigations are warranted to demonstrate the clinical relevance of DLR on low-dose CT in the near future. There were several limitations to this study. First, the study cohort was limited, and the number of otologic disease patients was small. These limitations are likely to have affected our study results. Second, both reconstruction methods for HDCT were provided by a single vendor and not compared with those provided by other vendors. Moreover, we used only the HDCT system, not other CT systems such as ADCT and PCDCT, while other reconstruction methods such as MBIR and conventional DLR were not used either in this study. Therefore, this fact is affected by our study results, and the study results were not transferrable to other CT scanners provided not only by the same but also other vendors. In addition, the reconstruction matrix in this study was a 512 × 512 matrix and not assessed CT images with a 1024 × 1024 matrix and a 2048 × 2048 matrix because of larger data storage and limited clinical availability. Furthermore, a few investigators have suggested that PCDCT can be used for middle and inner ears as well as temporal bones. These considerations must thus also have affected our study results. Third, we did not directly compare its effect on decision-making for patients’ treatment. The clinical relevance of our findings for patients’ care has therefore not been determined, and further studies using a large prospective cohort with otologic diseases are warranted. Fourth, this study did not use an in vitro study, and detailed analyses for spatial resolution improvements were not analyzed in detail, while further investigations with phantoms are also warranted. In conclusion, DLR showed superior potential to that of hybrid-type IR for better image quality and visualization of anatomical landmarks in middle and inner ears and temporal bones on HDCT, although diagnostic performance was not affected in clinical practice. Author Contributions In this study, all authors contributed as follows: conceptualization, M.N., T.Y., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); methodology, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); software, M.N., H.K., Y.I., K.F., N.A., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); validation, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno).; formal analysis, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); investigation, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); resources, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); data curation, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); writing—original draft preparation, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); writing—review and editing, M.N., H.K., Y.I., K.F., N.A., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); visualization, M.N., H.K., Y.I., K.F., N.A., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); supervision, Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno); project administration, M.N., T.U., T.Y., D.T., Y.O. (Yoshiyuki Ozawa) and Y.O. (Yoshiharu Ohno).; funding acquisition, Y.O. (Yoshiharu Ohno). All authors have read and agreed to the published version of the manuscript. Funding This study was financially and technically supported by Canon Medical Systems Corporation. Institutional Review Board Statement The retrospective study was conducted in accordance with the Declaration of Helsinki. Moreover, this study was approved by the Institutional Review Board of Fujita Health University Hospital, was compliant with the Health Insurance Portability and Accountability Act, and written informed consent was waived. This study was approved by our ethics committee and numbered as HM-20-603 on 21 March 2021, in our institution. Informed Consent Statement Patient consent was waived due to a retrospective study from routine clinical practice data. Data Availability Statement The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author. Acknowledgments The authors wish to thank Ryota Matsumoto, RT, Akio Katagata, RT, Yumi Kataoka, RT, Shigenobu Maeda, RT (Department of Radiology, Fujita Health University Hospital), and Shigeki Kobayashi, (Faculty of Radiological Technology, Fujita Health University School of Health Sciences) for their input and contribution to this study. Conflicts of Interest Masahiko Nomura and Yoshiharu Ohno received a research grant from Canon Medical Systems Corporation, which also supported this work financially and technically. Hirona Kimata, Yuya Ito, Kenji Fujii and Naruomi Akino are employees of Canon Medical Systems Corporation. They did not have control over any of the data used in this study. Takahiro Ueda, Takeshi Yoshikawa, Daisuke Takenaka and Yoshiyuki Ozawa had nothing to disclose. All authors and Yoshiharu Ohno, who is the last and corresponding author, have agreed to the publication of this study. All data were controlled by Masahiko Nomura, Takahiro Ueda, Takeshi Yoshikawa, Daisuke Takenaka, Yoshiyuki Ozawa and Yoshiharu Ohno. Moreover, all statistical analyses were performed by the same authors. References Zhou, W.; Lane, J.; Carlson, M.; Bruesewitz, M.; Witte, R.; Koeller, K.; Eckel, L.; Carter, R.; McCollough, C.; Leng, S. Comparison of a Photon-Counting-Detector CT with an Energy-Integrating-Detector CT for Temporal Bone Imaging: A Cadaveric Study. AJNR Am. J. Neuroradiol. 2018, 39, 1733–1738. [] [ CrossRef] Kurokawa, R.; Maeda, E.; Mori, H.; Amemiya, S.; Sato, J.; Ino, K.; Torigoe, R.; Abe, O. Evaluation of the depiction ability of the microanatomy of the temporal bone in quarter-detector CT: Model-based iterative reconstruction vs hybrid iterative reconstruction. Medicine 2019, 98, e15991. [] [ CrossRef] Ohara, A.; Machida, H.; Shiga, H.; Yamamura, W.; Yokoyama, K. Improved image quality of temporal bone CT with an ultrahigh-resolution CT scanner: Clinical pilot studies. Jpn. J. Radiol. 2020, 38, 878–883. [] [ CrossRef] [ PubMed] Heutink, F.; Klabbers, T.M.; Huinck, W.J.; Lucev, F.; van der Woude, W.J.; Mylanus, E.A.M.; Verbist, B.M. Ultra-High-Resolution CT to Detect Intracochlear New Bone Formation after Cochlear Implantation. Radiology 2022, 302, 605–612. [] [ CrossRef] Ide, S.; Urikura, A.; Yoshida, T.; Nakaya, Y.; Endo, M.; Aramaki, T. Ultrahigh-Resolution Computed Tomography Improves Preoperative Computed Tomography Angiography for Deep Inferior Epigastric Artery Perforator Flap Reconstruction. J. Comput. Assist. Tomogr. 2022, 46, 29–33. [] [ CrossRef] Grunz, J.-P.; Heidenreich, J.F.; Lennartz, S.; Weighardt, J.P.; Bley, T.A.; Ergün, S.; Petritsch, B.; Huflage, H. Spectral Shaping Via Tin Prefiltration in Ultra-High-Resolution Photon-Counting and Energy-Integrating Detector CT of the Temporal Bone. Investig. Radiol. 2022, 57, 819–825. [] [ CrossRef] Ohno, Y.; Takenaka, D.; Kanda, T.; Yoshikawa, T.; Matsumoto, S.; Sugihara, N.; Sugimura, K. Adaptive iterative dose reduction using 3D processing for reduced- and low-dose pulmonary CT: Comparison with standard-dose CT for image noise reduction and radiological findings. AJR Am. J. Roentgenol. 2012, 199, W477–W485. [] [ CrossRef] Gomez-Cardona, D.; Nagle, S.K.; Li, K.; Robinson, T.E.; Chen, G.H. Influence of radiation dose and reconstruction algorithm in MDCT assessment of airway wall thickness: A phantom study. Med. Phys. 2015, 42, 5919–5927. [] [ CrossRef] [ PubMed] Padole, A.; Singh, S.; Ackman, J.B.; Wu, C.; Do, S.; Pourjabbar, S.; Khawaja, R.D.A.; Otrakji, A.; Digumarthy, S.; Shepard, J.-A.; et al. Submillisievert chest CT with filtered back projection and iterative reconstruction techniques. AJR Am. J. Roentgenol. 2014, 203, 772–781. [] [ CrossRef] Mirsadraee, S.; Weir, N.; Connolly, S.; Murchison, J.; Reid, J.; Hirani, N.; Connell, M.; van Beek, E. Feasibility of radiation dose reduction using AIDR-3D in dynamic pulmonary CT perfusion. Clin. Radiol. 2015, 70, 844–851. [] [ CrossRef] [ PubMed] Wielpütz, M.O.; Wroblewski, J.; Lederlin, M.; Dinkel, J.; Eichinger, M.; Koenigkam-Santos, M.; Biederer, J.; Kauczor, H.-U.; Puderbach, M.U.; Jobst, B.J. Computer-aided detection of artificial pulmonary nodules using an ex vivo lung phantom: Influence of exposure parameters and iterative reconstruction. Eur. J. Radiol. 2015, 84, 1005–1011. [] [ CrossRef] Kim, C.; Lee, K.Y.; Shin, C.; Kang, E.-Y.; Oh, Y.-W.; Ha, M.; Ko, C.S.; Cha, J. Comparison of Filtered Back Projection, Hybrid Iterative Reconstruction, Model-Based Iterative Reconstruction, and Virtual Monoenergetic Reconstruction Images at Both Low- and Standard-Dose Settings in Measurement of Emphysema Volume and Airway Wall Thickness: A CT Phantom Study. Korean J. Radiol. 2018, 19, 809–817. [] Kim, D.J.; Kim, C.; Shin, C.; Lee, S.K.; Ko, C.S.; Lee, K.Y. Impact of Model-Based Iterative Reconstruction on the Correlation between Computed Tomography Quantification of a Low Lung Attenuation Area and Airway Measurements and Pulmonary Function Test Results in Normal Subjects. Korean J. Radiol. 2018, 19, 1187–1195. [] [ CrossRef] Matsukiyo, R.; Ohno, Y.; Matsuyama, T.; Nagata, H.; Kimata, H.; Ito, Y.; Ogawa, Y.; Murayama, K.; Kato, R.; Toyama, H. Deep learning-based and hybrid-type iterative reconstructions for CT: Comparison of capability for quantitative and qualitative image quality improvements and small vessel evaluation at dynamic CE-abdominal CT with ultra-high and standard resolutions. Jpn. J. Radiol. 2021, 39, 186–197. [] [ CrossRef] Hamabuchi, N.; Ohno, Y.; Kimata, H.; Ito, Y.; Fujii, K.; Akino, N.; Takenaka, D.; Yoshikawa, T.; Oshima, Y.; Matsuyama, T.; et al. Effectiveness of deep learning reconstruction on standard to ultra-low-dose high-definition chest CT images. Jpn. J. Radiol. 2023, 41, 1373–1388. [] [ CrossRef] Tsuchiya, N.; Kobayashi, S.; Nakachi, R.; Tomori, Y.; Yogi, A.; Iida, G.; Ito, J.; Nishie, A. Application of a pulmonary nodule detection program using AI technology to ultra-low-dose CT: Differences in detection ability among various image reconstruction methods. Jpn. J. Radiol. 2025, 43, 1303–1312. [] [ CrossRef] [ PubMed] Willemink, M.J.; Noël, P.B. The evolution of image reconstruction for CT-from filtered back projection to artificial intelligence. Eur. Radiol. 2019, 29, 2185–2195. [] [ CrossRef] [ PubMed] Koetzier, L.R.; Mastrodicasa, D.; Szczykutowicz, T.P.; van der Werf, N.R.; Wang, A.S.; Sandfort, V.; van der Molen, A.J.; Fleischmann, D.; Willemink, M.J. Deep Learning Image Reconstruction for CT: Technical Principles and Clinical Prospects. Radiology 2023, 306, e221257. [] [ CrossRef] [ PubMed] Danish Society of Radiology. European Guidelines on Quality Criteria for Computed Tomography. 1998. Available online: https://op.europa.eu/en/publication-detail/-/publication/d229c9e1-a967-49de-b169-59ee68605f1a (accessed on 8 January 2025). Meyer, M.; Haubenreisser, H.; Raupach, R.; Schmidt, B.; Lietzmann, F.; Leidecker, C.; Allmendinger, T.; Flohr, T.; Schad, L.R.; Schoenberg, S.O.; et al. Initial results of a new generation dual source CT system using only an in-plane comb filter for ultra-high resolution temporal bone imaging. Eur. Radiol. 2015, 25, 178–185. [] [ CrossRef] Shinagare, A.B.; Alper, D.P.; Hashemi, S.R.; Chai, J.L.; Hammer, M.M.; Boland, G.W.; Khorasani, R. Early adoption of a certainty scale to improve diagnostic certainty communication. J. Am. Coll. Radiol. 2020, 17, 1276–1284. [] [ CrossRef] Shinagare, A.B.; Shankar, P.R.; Chernyak, V.; Woolen, S.A.; Herts, B.R.; Azene, E.M.; Mitchell, D.G.; Rosenkrantz, A.B.; Raghavan, K.; Karmazyn, B.; et al. Potential Frameworks for Communicating Diagnostic Certainty in Radiology Reports: From the ACR Commission on Quality and Safety. J. Am. Coll. Radiol. 2025, 22, 1390–1398. [] [ CrossRef] Svanholm, H.; Starklint, H.; Gundersen, H.J.; Fabricius, J.; Barlebo, H.; Olsen, S. Reproducibility of histomorphologic diagnoses with special reference to the kappa statistic. APMIS 1989, 97, 689–698. [] [ CrossRef] Kanda, Y. Investigation of the freely available easy-to-use software ’EZR’ for medical statistics. Bone Marrow Transpl. 2013, 48, 452–458. [] [ CrossRef] Faul, F.; Erdfelder, E.; Lang, A.G.; Buchner, A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav. Res. Methods 2007, 39, 175–191. [] [ CrossRef] Figure 1. Patients’ selection flowchart for this study. Figure 1. Patients’ selection flowchart for this study. Figure 2. HDCT images at the inner ear level reconstructed by means of hybrid-type IR and DLR methods. ( Left: hybrid IR, Right: DLR). Significant image noise and artifact reductions and improvements in SNR and overall image quality were confirmed by comparing HDCT images reconstructed with hybrid-type IR and with DLR, although the visibility of anatomical landmarks had little effect. Figure 2. HDCT images at the inner ear level reconstructed by means of hybrid-type IR and DLR methods. ( Left: hybrid IR, Right: DLR). Significant image noise and artifact reductions and improvements in SNR and overall image quality were confirmed by comparing HDCT images reconstructed with hybrid-type IR and with DLR, although the visibility of anatomical landmarks had little effect. Figure 3. HDCT images at the middle ear level reconstructed by means of hybrid-type IR and DLR methods. ( Left: hybrid IR, Right: DLR). Significant image noise and artifact reductions and improvements in SNR and overall image quality were confirmed by comparing HDCT images reconstructed with hybrid-type IR and with DLR, although the visibility of anatomical landmarks had little effect. Figure 3. HDCT images at the middle ear level reconstructed by means of hybrid-type IR and DLR methods. ( Left: hybrid IR, Right: DLR). Significant image noise and artifact reductions and improvements in SNR and overall image quality were confirmed by comparing HDCT images reconstructed with hybrid-type IR and with DLR, although the visibility of anatomical landmarks had little effect. Figure 4. HDCT images in an otosclerosis patient at the inner ear level reconstructed by means of hybrid-type IR and DLR methods. ( Left: hybrid IR, Right: DLR). Significant image noise and artifact reductions and improvements in SNR and overall image quality were confirmed by comparing HDCT images reconstructed with hybrid-type IR and with DLR. HDCT image reconstructed with DLR demonstrates incus synostosis and calcification of the stapedius tendon more clearly than that with hybrid-type IR. The diagnostic confidence level of the HDCT image reconstructed with DLR was 5, although that with hybrid-type IR was 4. Figure 4. HDCT images in an otosclerosis patient at the inner ear level reconstructed by means of hybrid-type IR and DLR methods. ( Left: hybrid IR, Right: DLR). Significant image noise and artifact reductions and improvements in SNR and overall image quality were confirmed by comparing HDCT images reconstructed with hybrid-type IR and with DLR. HDCT image reconstructed with DLR demonstrates incus synostosis and calcification of the stapedius tendon more clearly than that with hybrid-type IR. The diagnostic confidence level of the HDCT image reconstructed with DLR was 5, although that with hybrid-type IR was 4. Figure 5. Compared results of ROC analysis between HDCTs reconstructed with hybrid-type IR and DLR methods (blue line: hybrid-type IR; red line: DLR). There was no significant difference in AUC between the two reconstruction methods (DLR: AUC = 0.996, hybrid-type IR: AUC = 0.993, p = 0.18). Figure 5. Compared results of ROC analysis between HDCTs reconstructed with hybrid-type IR and DLR methods (blue line: hybrid-type IR; red line: DLR). There was no significant difference in AUC between the two reconstruction methods (DLR: AUC = 0.996, hybrid-type IR: AUC = 0.993, p = 0.18). Table 1. Patients’ characteristics. Table 1. Patients’ characteristics. Gender (Cases) Male 69 Female 71 Age (years) All cases Mean 61 Range 11–86 Male Mean 62 Range 18–84 Female Mean 60 Range 11–86 Diagnosis (cases) Without diseases 108 With diseases 32 Underlying diseases Otitis media cholesteatoma 9 Chronic otitis media 7 Ossicular chain disruption 6 Otosclerosis 5 temporal bone fracture 5 Table 2. Inter-observer agreements for overall image quality and artifacts. Table 2. Inter-observer agreements for overall image quality and artifacts. Qualitative Index Reconstruction Method Observers Visual Score κ p Value 1 2 3 4 5 Overall image quality DLR 1 0 0 0 22 194 0.61 <0.0001 2 0 0 0 44 172 hybrid-type IR 1 0 0 0 98 118 0.63 <0.0001 2 0 0 10 120 86 Artifacts DLR 1 194 22 0 0 0 1.0 <0.0001 2 194 22 0 0 0 hybrid-type IR 1 118 98 0 0 0 0.63 <0.0001 2 108 98 10 0 0 DLR: deep learning reconstruction, IR: iterative reconstruction. Table 3. Inter-observer agreements for detailed evaluation of the visibility of anatomical landmarks in middle and inner ears and temporal bone. Table 3. Inter-observer agreements for detailed evaluation of the visibility of anatomical landmarks in middle and inner ears and temporal bone. Anatomical Structure Reconstruction Method Observers Visual Score κ p Value 1 2 3 4 5 Middle ear DLR 1 0 0 0 33 183 0.80 <0.0001 2 0 0 0 44 172 hybrid-type IR 1 0 0 10 120 86 0.62 <0.0001 2 0 0 10 121 85 Inner ear DLR 1 0 0 0 22 194 0.62 <0.0001 2 0 0 0 43 173 hybrid-type IR 1 0 0 11 87 118 0.73 <0.0001 2 0 0 11 119 86 Temporal bone DLR 1 0 0 0 32 184 0.62 <0.0001 2 0 0 0 33 183 hybrid-type IR 1 0 0 0 119 97 0.62 <0.0001 2 0 0 10 121 85 DLR: deep learning reconstruction, IR: iterative reconstruction. Table 4. Compared diagnostic performance between two reconstruction methods. Table 4. Compared diagnostic performance between two reconstruction methods. Reconstruction Method DLR Hybrid-Type IR p Value Area under the curve (AUC) 0.996 0.993 0.18 Sensitivity (%) 93.8 93.8 1.00 (30/32) (30/32) Specificity (%) 100 98.1 0.50 (108/108) (106/108) Positive predictive value (%) 100 98.1 N/A (30/30) (30/32) Negative predictive value (%) 98.2 98.1 N/A (108/110) (106/108) Accuracy (%) 98.6 97.1 0.50 (138/140) (136/140) DLR: deep learning reconstruction, IR: iterative reconstruction, N/A: not applicable. Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. © 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license. Share and Cite Nomura, M.; Kimata, H.; Ito, Y.; Fujii, K.; Akino, N.; Ueda, T.; Yoshikawa, T.; Takenaka, D.; Ozawa, Y.; Ohno, Y. Deep Learning Reconstruction Specialized for Inner Ear: Improving Image Quality and Anatomical Structure Visualization as Compared with Conventional Hybrid-Type Iterative Reconstruction on High-Definition CT. Diagnostics 2026, 16, 1756. https://doi.org/10.3390/diagnostics16121756 Nomura M, Kimata H, Ito Y, Fujii K, Akino N, Ueda T, Yoshikawa T, Takenaka D, Ozawa Y, Ohno Y. Deep Learning Reconstruction Specialized for Inner Ear: Improving Image Quality and Anatomical Structure Visualization as Compared with Conventional Hybrid-Type Iterative Reconstruction on High-Definition CT. Diagnostics. 2026; 16(12):1756. https://doi.org/10.3390/diagnostics16121756 Nomura, Masahiko, Hirona Kimata, Yuya Ito, Kenji Fujii, Naruomi Akino, Takahiro Ueda, Takeshi Yoshikawa, Daisuke Takenaka, Yoshiyuki Ozawa, and Yoshiharu Ohno. 2026. "Deep Learning Reconstruction Specialized for Inner Ear: Improving Image Quality and Anatomical Structure Visualization as Compared with Conventional Hybrid-Type Iterative Reconstruction on High-Definition CT" Diagnostics 16, no. 12: 1756. https://doi.org/10.3390/diagnostics16121756 Nomura, M., Kimata, H., Ito, Y., Fujii, K., Akino, N., Ueda, T., Yoshikawa, T., Takenaka, D., Ozawa, Y., & Ohno, Y. (2026). Deep Learning Reconstruction Specialized for Inner Ear: Improving Image Quality and Anatomical Structure Visualization as Compared with Conventional Hybrid-Type Iterative Reconstruction on High-Definition CT. Diagnostics, 16(12), 1756. https://doi.org/10.3390/diagnostics16121756 Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details . Article Metrics Article metric data becomes available approximately 24 hours after publication online.
