EUVIMEDEUVIMED entdecken
Open Access

Fast Submillimeter Whole‐Brain T 2 *‐Weighted Imaging Using 3D ‐ EPI With CAIPIRINHA and Deep‐Learning Denoising at 3T

Sreekanth Madhusoodhanan Nair iD, Bryan Quah, Jae W Song, Jin Jin iD, Chang Gao, Fei Han, Elaina Luskin, Michael Luu, Brian Renner, Ke Cheng Liu, Sunil Patil, John A Derbyshire und 17 weitere

Magnetic Resonance in Medicine · 2026

DOI: 10.1002/mrm.70522 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Magnetic Resonance in Medicine, 2026. Verfasst von Sreekanth Madhusoodhanan Nair und 28 weiteren Autor:innen. Beteiligte Einrichtung: Neuroimaging Program, Department of Neurology, Cedars-Sinai Medical Center, Los Angeles, California, USA. Hauptthemen laut MEDLINE: Imaging, Three-Dimensional, Echo-Planar Imaging, Deep Learning, Brain. Die Arbeit nennt 47 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NINDS NIH HHS, NIMH NIH HHS, the Cedars-Sinai Women's Guild Neurology Project.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Recent updates to the diagnostic criteria of multiple sclerosis (MS) require whole‐brain T 2 *‐weighted (T 2 *w) imaging with submillimeter resolution to detect novel diagnostic biomarkers such as the central vein sign. However, to achieve the needed submillimeter spatial resolution, conventional T 2 *w 3D gradient‐echo scans sequences are limited by prohibitively long scan times for clinical use. Here, we evaluated a different approach based on a segmented 3D echo planar imaging (3D‐EPI) sequence, accelerated with 2D Controlled Aliasing in Parallel Imaging Results in Higher Acceleration (CAIPIRINHA) undersampling and denoised with a deep learning‐based network.

Methods

Fifty‐two research participants were imaged at 3T using the 3D‐EPI sequence acquired at different CAIPIRINHA acceleration factors ( R = 2, 3, and 4) and denoised using a dedicated denoising convolutional neural network (DnCNN). Quantitative assessment of the accelerated T 2 *w 3D‐EPI scans, before and after denoising, was performed using peak signal‐to‐noise ratio (PSNR), structural similarity index (SSIM), and tissue contrasts. A neuroradiologist separately assessed image quality in a blinded manner using predetermined scoring criteria.

Results

T 2 *w 3D‐EPI with CAIPIRINHA acceleration enabled fast submillimeter isotropic (0.65 mm) imaging of the entire brain with scan times ranging between 3 min 22 s ( R = 2) down to 1 min 56 s ( R = 4). Even for the fastest scan ( R = 4), accelerated T 2 *w 3D‐EPI images denoised with DnCNN exhibited superior PSNR (3 dB increase), SSIM (13% increase) and lesion‐to‐vein tissue contrast (9% increase) compared to the non‐denoised images.

Conclusions

The 3D‐EPI sequence combined with CAIPIRINHA and deep learning denoising enables fast submillimeter whole‐brain T 2 *w imaging at 3T.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Imaging, Three-DimensionalEcho-Planar ImagingDeep LearningBrainMultiple SclerosisHumansSignal-To-Noise RatioFemaleMaleAdultConvolutional Neural NetworksAlgorithmsImage Interpretation, Computer-AssistedImage Processing, Computer-AssistedMiddle AgedImage Enhancement

Stichwörter der Autor:innen

3D‐EPICAIPIRINHA accelerationdeep learning denoisingsubmillimeter resolution

Autor:innen und Institutionen

  • Neuroimaging Program, Department of Neurology, Cedars-Sinai Medical Center, Los Angeles, California, USA.
  • Department of Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, USA.
  • Siemens Healthineers Pty Ltd, Brisbane, Australia.
  • Siemens Healthineers, Malvern, Pennsylvania, USA.
  • Biostatistics Shared Resource, Department of Computational Biomedicine, Cedars-Sinai Medical Center, Los Angeles, California, USA.
  • Functional MRI Facility, National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland, USA.

ORCID-Profile: Sreekanth Madhusoodhanan Nair, Jin Jin, Matthew K Schindler

Förderung

  • NINDS NIH HHS
  • NIMH NIH HHS
  • the Cedars-Sinai Women's Guild Neurology Project
  • the Erwin Rautenberg Foundation
  • the National MS Society (NMSS)
  • National Institute of Neurological Disorders and Stroke
  • National Multiple Sclerosis Society

Zitieren

APA

Nair, S. M., Quah, B., Song, J. W., Jin, J., Gao, C., Han, F., Luskin, E., Luu, M., Renner, B., Liu, K. C., Patil, S., Derbyshire, J. A., Sakaie, K., Lee, J., Elliott, M. A., Shinohara, R. T., Schindler, M. K., Bilello, M., Kaisey, M., …, & Sati, P. (2026). Fast Submillimeter Whole‐Brain T 2 *‐Weighted Imaging Using 3D ‐ EPI With CAIPIRINHA and Deep‐Learning Denoising at 3T. Magnetic Resonance in Medicine. https://doi.org/10.1002/mrm.70522

Vancouver

Nair SM, Quah B, Song JW, Jin J, Gao C, Han F, et al. Fast Submillimeter Whole‐Brain T 2 *‐Weighted Imaging Using 3D ‐ EPI With CAIPIRINHA and Deep‐Learning Denoising at 3T. Magnetic Resonance in Medicine. 2026. doi:10.1002/mrm.70522

BibTeX
@article{Nair2026,
  author = {Nair, Sreekanth Madhusoodhanan and Quah, Bryan and Song, Jae W and Jin, Jin and Gao, Chang and Han, Fei and Luskin, Elaina and Luu, Michael and Renner, Brian and Liu, Ke Cheng and Patil, Sunil and Derbyshire, John A and Sakaie, Ken and Lee, Jonathan and Elliott, Mark A and Shinohara, Russell T and Schindler, Matthew K and Bilello, Michel and Kaisey, Marwa and Al-Louzi, Omar and Binesh, Nader and Maya, Marcel and Li, Debiao and Solomon, Andrew J and Reich, Daniel S and Sicotte, Nancy L and Lowe, Mark J and Ontaneda, Daniel and Sati, Pascal},
  title = {Fast Submillimeter Whole‐Brain T 2 *‐Weighted Imaging Using 3D ‐ EPI With CAIPIRINHA and Deep‐Learning Denoising at 3T},
  journal = {Magnetic Resonance in Medicine},
  year = {2026},
  doi = {10.1002/mrm.70522}
}
Kostenlos starten →