Skip to content
2000
Volume 21, Issue 1
  • ISSN: 1573-4056
  • E-ISSN: 1875-6603
file format text download EPUB
file format pdf download PDF
side by side viewer icon HTML

Abstract

Purposes

Deep learning methods were employed to perform harmonization analysis on whole-brain scans obtained from 1.5-T and 3.0-T scanners, aiming to increase comparability between different magnetic resonance imaging (MRI) scanners.

Methods

Thirty patients evaluated in Beijing Tsinghua Changgung Hospital between August 2020 and March 2023 were included in this retrospective study. Three MRI scanners were used to scan patients, and automated brain image segmentation was performed to obtain volumes of different brain regions. Differences in regional volumes across scanners were analyzed using repeated-measures analysis of variance. For regions showing significant differences, super-resolution deep learning was applied to enhance consistency, with subsequent comparison of results. For regions still exhibiting differences, the Intraclass Correlation Coefficient (ICC) was calculated and the consistency was evaluated using Cicchetti's criteria.

Results

Average whole-brain volumes for different scanners among patients were 1152.36mm3 (SD = 95.34), 1136.92mm3 (SD = 108.21), and 1184.00mm3 (SD = 102.78), respectively. Analysis revealed significant variations in all 12 brain regions (p<0.05), indicating a lack of comparability among imaging results obtained from different magnetic field strengths. After deep learning-based consistency optimization, most brain regions showed no significant differences, except for six regions where differences remained significant. Among these, three regions demonstrated ICC values of 0.868 (95%CI 0.771-0.931), 0.776 (95%CI 0.634-0.877), and 0.893 (95%CI 0.790-0.947), indicating high reproducibility and comparability.

Discussion

This study demonstrates a deep learning-based harmonization method that effectively mitigates field strength-related inconsistencies between 1.5-T and 3.0-T MRI, significantly enhancing their comparability. The high ICCs observed in key brain regions confirm the robustness of this approach, paving the way for reliable clinical application across different scanners. A noted limitation is its current focus on brain imaging, which warrants future research to extend its applicability to other anatomical areas.

Conclusion

This study employed a novel machine learning approach that significantly improved the comparability of imaging results from patients using different magnetic field strengths and various models of MRI scanners. Furthermore, it enhanced the consistency of central nervous system image segmentation.

© 2025 The Author(s). Published by Bentham Science Publishers.
This is an open access article published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
Loading

Article metrics loading...

/content/journals/cmir/10.2174/0115734056383931250919073319
2025-10-02
2025-11-08

  • From This Site

    /content/journals/cmir/10.2174/0115734056383931250919073319
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

/deliver/fulltext/cmir/21/1/CMIR-21-E15734056383931.html?itemId=/content/journals/cmir/10.2174/0115734056383931250919073319&mimeType=html&fmt=ahah

References

  1. SahaD.K. BohsaliA. SahaR. HajjarI. CalhounV.D. Neuromark PET: A multivariate method for estimating and comparing whole brain functional networks and connectomes from fMRI and PET data.bioRxiv202410.1101/2024.01.10.575131
     [Google Scholar]
  2. MarquesJ.P. SimonisF.F.J. WebbA.G. Low‐field MRI: An MR physics perspective.J Magn Reson Imaging20194961528154210.1002/jmri.2663730637943
     [Google Scholar]
  3. LiaoB. ChenY. WangZ. SmithC. LiuJ. A comparative study on 1.5T–3T MRI conversion through deep neural network models.arXiv202210.48550/arXiv.2210.06362
     [Google Scholar]
  4. VictoriaT. JohnsonA.M. EdgarJ.C. ZarnowD.M. VossoughA. JaramilloD. Comparison between 1.5-T and 3-T MRI for fetal imaging: is there an advantage to imaging with a higher field strength?.AJR Am J Roentgenol2016206119520110.2214/AJR.14.1420526700352
     [Google Scholar]
  5. DeLanoM.C. FisherC. 3T MR imaging of the brain.Magn Reson Imaging Clin N Am2006141778810.1016/j.mric.2006.01.00416530636
     [Google Scholar]
  6. VictoriaT. JaramilloD. RobertsT.P.L. ZarnowD. JohnsonA.M. DelgadoJ. RubesovaE. VossoughA. Fetal magnetic resonance imaging: Jumping from 1.5 to 3 tesla (preliminary experience).Pediatr Radiol201444437638610.1007/s00247‑013‑2857‑024671739
     [Google Scholar]
  7. BarreraC.A. FrancavillaM.L. SeraiS.D. EdgarJ.C. JaimesC. GeeM.S. RobertsT.P.L. OteroH.J. AdzickN.S. VictoriaT. Specific absorption rate and specific energy dose: comparison of 1.5-T versus 3.0-T fetal MRI.Radiology2020295366467410.1148/radiol.202019155032255418
     [Google Scholar]
  8. MinhasA.S. OliverR. Magnetic resonance imaging basics.Adv Exp Med Biol20221380478210.1007/978‑3‑031‑03873‑0_336306094
     [Google Scholar]
  9. RuttB.K. LeeD.H. The impact of field strength on image quality in MRI.J Magn Reson Imaging199661576210.1002/jmri.18800601118851404
     [Google Scholar]
  10. HoltackersR.J. WildbergerJ.E. WinterspergerB.J. ChiribiriA. Impact of field strength in clinical cardiac magnetic resonance imaging.Invest Radiol2021561176477210.1097/RLI.000000000000080934261084
     [Google Scholar]
  11. Astivia-ChávezD.N. RodríguezA.O. Ortiz-PosadasM.R. Performance evaluation of a magnetic resonance imager with evidence-based decision making tools.Annu Int Conf IEEE Eng Med Biol Soc202320231410.1109/EMBC40787.2023.1034040138083537
     [Google Scholar]
  12. XianJ.F. ChenM. JinZ.Y. Magnetic resonance imaging in clinical medicine: Current status and potential future developments in China.Chin Med J2015128556957010.4103/0366‑6999.15163725698184
     [Google Scholar]
  13. MavrogeniS.I. PoulosG. KolovouG. TheodorakisG. Magnetic resonance imaging-conditional devices: Luxury or real clinical need?Hellenic J Cardiol201758425626010.1016/j.hjc.2017.01.00128089649
     [Google Scholar]
  14. SuttonR. KanalE. WilkoffB.L. BelloD. LuechingerR. JenniskensI. HullM. SommerT. Safety of magnetic resonance imaging of patients with a new Medtronic EnRhythm MRI SureScan pacing system: clinical study design.Trials2008916810.1186/1745‑6215‑9‑6819055703
     [Google Scholar]
  15. MorraJ.H. TuZ. ApostolovaL.G. GreenA.E. AvedissianC. MadsenS.K. ParikshakN. HuaX. TogaA.W. JackC.R. SchuffN. WeinerM.W. ThompsonP.M. Automated 3D mapping of hippocampal atrophy and its clinical correlates in 400 subjects with Alzheimer’s disease, mild cognitive impairment, and elderly controls.Hum Brain Mapp20093092766278810.1002/hbm.2070819172649Alzheimer’s Disease Neuroimaging Initiative
     [Google Scholar]
  16. ApostolovaL.G. ThompsonP.M. GreenA.E. HwangK.S. ZoumalanC. JackC.R. HarveyD.J. PetersenR.C. ThalL.J. AisenP.S. TogaA.W. CummingsJ.L. DeCarliC.S. 3D comparison of low, intermediate, and advanced hippocampal atrophy in MCI.Hum Brain Mapp201031578679710.1002/hbm.2090520143386
     [Google Scholar]
  17. HuaX. LeowA. LeeS. KlunderA. TogaA. LeporeN. ChouY. BrunC. ChiangM. BaryshevaM. JackC.R. BernsteinM.A. BritsonP.J. WardC.P. WhitwellJ.L. BorowskiB. FleisherA.S. FoxN.C. BoyesR.G. BarnesJ. HarveyD. KornakJ. SchuffN. BoretaL. AlexanderG.E. WeinerM.W. ThompsonP.M. 3D characterization of brain atrophy in Alzheimer’s disease and mild cognitive impairment using tensor-based morphometry.Neuroimage2008411193410.1016/j.neuroimage.2008.02.01018378167Alzheimer’s Disease Neuroimaging Initiative
     [Google Scholar]
  18. JackC.R. BernsteinM.A. FoxN.C. ThompsonP. AlexanderG. HarveyD. BorowskiB. BritsonP.J. L WhitwellJ. WardC. DaleA.M. FelmleeJ.P. GunterJ.L. HillD.L. KillianyR. SchuffN. Fox-BosettiS. LinC. StudholmeC. DeCarliC.S. KruegerG. WardH.A. MetzgerG.J. ScottK.T. MallozziR. BlezekD. LevyJ. DebbinsJ.P. FleisherA.S. AlbertM. GreenR. BartzokisG. GloverG. MuglerJ. WeinerM.W. The Alzheimer’s Disease Neuroimaging Initiative (ADNI): MRI methods.J Magn Reson Imaging200827468569110.1002/jmri.2104918302232
     [Google Scholar]
  19. KooT.K. LiM.Y. A guideline of selecting and reporting intraclass correlation coefficients for reliability research.J Chiropr Med201615215516310.1016/j.jcm.2016.02.01227330520
     [Google Scholar]
  20. ZhangY. BradyM. SmithS. Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm.IEEE Trans Med Imaging2001201455710.1109/42.90642411293691
     [Google Scholar]
  21. WeiJ. The adoption of repeated measurement of variance analysis and Shapiro—Wilk test.Front Med202216465966010.1007/s11684‑021‑0908‑835776406
     [Google Scholar]
  22. ZhaoX. WenZ. Super-resolution of diffusion-weighted images using space-customized learning model.Technol Health Care202432S142343510.3233/THC‑24803738759065
     [Google Scholar]
  23. ChengC.H. HuangH.M. LinH.L. ChiouS.M. 1.5T versus 3T MRI for targeting subthalamic nucleus for deep brain stimulation.Br J Neurosurg201428446747010.3109/02688697.2013.85431224191703
     [Google Scholar]
  24. YasakaK. TanishimaT. OhtakeY. TajimaT. AkaiH. OhtomoK. AbeO. KiryuS. Deep learning reconstruction for the evaluation of neuroforaminal stenosis using 1.5T cervical spine MRI: Comparison with 3T MRI without deep learning reconstruction.Neuroradiology202264102077208310.1007/s00234‑022‑03024‑635918450
     [Google Scholar]
  25. ChowN. HwangK.S. HurtzS. GreenA.E. SommeJ.H. ThompsonP.M. ElashoffD.A. JackC.R. WeinerM. ApostolovaL.G. Comparing 3T and 1.5T MRI for mapping hippocampal atrophy in the Alzheimer’s disease neuroimaging initiative.AJNR Am J Neuroradiol201536465366010.3174/ajnr.A422825614473Alzheimer’s Disease Neuroimaging Initiative
     [Google Scholar]
  26. NathV. RemediosS. ParvathaneniP. HansenC.B. BayrakR.G. BermudezC. BlaberJ.A. SchillingK.G. JanveV.A. GaoY. HuoY. LyuI. WilliamsO. ResnickS. Beason-HeldL. RogersB.P. StepniewskaI. AndersonA.W. LandmanB.A. Harmonizing 1.5T/3T diffusion weighted mri through development of deep learning stabilized microarchitecture estimators.<comment xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML">San Diego, California, United States, 15 March,2019,10949 </comment>Proc SPIE Int Soc Opt Eng
     [Google Scholar]
  27. ManiM. YangB. BathlaG. MagnottaV. JacobM. Multi‐band‐ and in‐plane‐accelerated diffusion MRI enabled by model‐based deep learning in q‐space and its extension to learning in the spherical harmonic domain.Magn Reson Med20228741799181510.1002/mrm.2909534825729
     [Google Scholar]
  28. KashiwagiN. TanakaH. YamashitaY. TakahashiH. KassaiY. FujiwaraM. TomiyamaN. Applicability of deep learning-based reconstruction trained by brain and knee 3T MRI to lumbar 1.5T MRI.Acta Radiol Open20211062058460121102393910.1177/2058460121102393934211738
     [Google Scholar]
  29. AthithanL. GulsinG.S. HouseM.J. PangW. BradyE.M. WormleightonJ. ParkeK.S. Graham-BrownM. St PierreT.G. LeveltE. McCannG.P. A comparison of liver fat fraction measurement on MRI at 3T and 1.5T.PLoS One2021167e025292810.1371/journal.pone.025292834255778
     [Google Scholar]
  30. StadlbauerA PrayerD. PrayerD. Fetal MRI at higher field strength.Fetal MRI Medical RadiologyBerlin, HeidelbergSpringer2010334710.1007/174_2010_27
     [Google Scholar]
  31. BernsteinM.A. HustonJ. WardH.A. Imaging artifacts at 3.0T.J Magn Reson Imaging200624473574610.1002/jmri.2069816958057
     [Google Scholar]
  32. MerkleE.M. DaleB.M. PaulsonE.K. Abdominal MR imaging at 3T.Magn Reson Imaging Clin N Am2006141172610.1016/j.mric.2005.12.00116530632
     [Google Scholar]
  33. ICNIRP statement on the “guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz)”.Health Phys200997325725810.1097/HP.0b013e3181aff9db19667809International Commission on Non-Ionizing Radiation Protection (ICNIRP)
     [Google Scholar]
  34. Commission IE. International standard, medical equipment-part 2: Particular requirements for the safety of magnetic resonance equipment for medical diagnosis, 2nd revision.<comment xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML">Available from: <ext-link ext-link-type="uri" xlink:href="https://cdn.standards.iteh.ai/samples/104367/9c0294eea03f4c19a44433f89fcce97a/IEC-60601-2-33-2022.pdf">https://cdn.standards.iteh.ai/samples/104367/9c0294eea03f4c19a44433f89fcce97a/IEC-60601-2-33-2022.pdf</ext-link></comment>2002
  35. LaddM.E. BachertP. MeyerspeerM. MoserE. NagelA.M. NorrisD.G. SchmitterS. SpeckO. StraubS. ZaissM. Pros and cons of ultra-high-field MRI/MRS for human application.Prog Nucl Magn Reson Spectrosc201810915010.1016/j.pnmrs.2018.06.00130527132
     [Google Scholar]
/content/journals/cmir/10.2174/0115734056383931250919073319
Loading
A Comparative Study of Consistency on 1.5-T to 3.0-T Magnetic Resonance Imaging Conversion
Curr. Med. Imaging 21, E15734056383931 (2025); https://doi.org/10.2174/0115734056383931250919073319
/content/journals/cmir/10.2174/0115734056383931250919073319
/content/journals/cmir/10.2174/0115734056383931250919073319
Loading

Data & Media loading...


  • Article Type:     Research Article
Keyword(s): Brain segmentation; Cicchetti's criteria; Interclass correlation coefficient; Magnetic resonance imaging; Repeated measures analysis of variance; Signal-to-noise ratio

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error
Please enter a valid_number test