EUVIMEDEUVIMED entdecken
Multizentrische StudieOpen Access · CC BY-NC 4.0

Automatic detection of arterial input function for brain DCE ‐ MRI in multi‐site cohorts

Lucas Saca iD, Raghav Gaggar, Ioannis Pappas, Tammie Benzinger, Eric M Reiman, Mark S Shiroishi, Elizabeth B Joe, John M Ringman, Hussein N Yassine, Lon S Schneider, Helena C Chui, Daniel A Nation und 4 weitere

Magnetic Resonance in Medicine · 2025-12 · 94 (6) · 2732-2744

DOI: 10.1002/mrm.70020 ↗

Kurz gefasst

Multizentrische Studie in der Zeitschrift Magnetic Resonance in Medicine, online erschienen am 13.08.2025. Verfasst von Lucas Saca und 15 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, Loma Linda University, Loma Linda, California, USA. Hauptthemen laut MEDLINE: Magnetic Resonance Imaging, Contrast Media, Brain, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: NIA NIH HHS, Loma Linda University GRASP, NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Arterial input function (AIF) extraction is a crucial step in quantitative pharmacokinetic modeling of DCE‐MRI. This work proposes a robust deep learning model that can precisely extract an AIF from DCE‐MRI images.

Methods

A diverse dataset of human brain DCE‐MRI images from 289 participants, totaling 384 scans, from five different institutions with extracted gadolinium‐based contrast agent curves from large penetrating arteries, and with most data collected for blood–brain barrier (BBB) permeability measurement, was retrospectively analyzed. A 3D UNet model was implemented and trained on manually drawn AIF regions. The testing cohort was compared using proposed AIF quality metric AIFitness and K trans values from a standard DCE pipeline. This UNet was then applied to a separate dataset of 326 participants with a total of 421 DCE‐MRI images with analyzed AIF quality and K trans values.

Results

The resulting 3D UNet model achieved an average AIFitness score of 93.9 compared to 99.7 for manually selected AIFs, and white matter K trans values were 0.45/min × 10 −3 and 0.45/min × 10 −3 , respectively. The intraclass correlation between automated and manual K trans values was 0.89. The separate replication dataset yielded an AIFitness score of 97.0 and white matter K trans of 0.44/min × 10 −3 .

Conclusion

Findings suggest a 3D UNet model with additional convolutional neural network kernels and a modified Huber loss function achieves superior performance for identifying AIF curves from DCE‐MRI in a diverse multi‐center cohort. AIFitness scores and DCE‐MRI‐derived metrics, such as K trans maps, showed no significant differences in gray and white matter between manually drawn and automated AIFs.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Magnetic Resonance ImagingContrast MediaBrainImage Processing, Computer-AssistedArteriesHumansMaleFemaleMiddle AgedRetrospective StudiesAdultDeep LearningAgedBlood-Brain BarrierAlgorithmsCohort StudiesImage Interpretation, Computer-AssistedGadoliniumDynamic Contrast Enhanced Magnetic Resonance Imaging

Stichwörter der Autor:innen

3D UNetAIF detectionDCE‐MRIartificial intelligenceblood–brain barriermulti‐site data

Substanzen

Contrast MediaGadolinium

Autor:innen und Institutionen

  • Department of Radiology, Loma Linda University, Loma Linda, California, USA.
  • Department of Physiology and Neuroscience, Keck School of Medicine, University of Southern California, Los Angeles, California, USA.
  • Zilkha Neurogenetic Institute, Keck School of Medicine, University of Southern California, Los Angeles, California, USA.
  • Laboratory of NeuroImaging, USC Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, California, USA.
  • Mallinckrodt Institute of Radiology, Washington University in St. Louis School of Medicine, St. Louis, Missouri, USA.
  • The Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, USA.

ORCID-Profile: Lucas Saca, Samuel Barnes

Förderung

  • NIA NIH HHS
  • Loma Linda University GRASP
  • NIH HHS
  • Loma Linda University
  • National Institutes of Health

Zitieren

APA

Saca, L., Gaggar, R., Pappas, I., Benzinger, T., Reiman, E. M., Shiroishi, M. S., Joe, E. B., Ringman, J. M., Yassine, H. N., Schneider, L. S., Chui, H. C., Nation, D. A., Zlokovic, B. V., Toga, A. W., Chakhoyan, A., & Barnes, S. (2025). Automatic detection of arterial input function for brain DCE ‐ MRI in multi‐site cohorts. Magnetic Resonance in Medicine, 94(6), 2732-2744. https://doi.org/10.1002/mrm.70020

Vancouver

Saca L, Gaggar R, Pappas I, Benzinger T, Reiman EM, Shiroishi MS, et al. Automatic detection of arterial input function for brain DCE ‐ MRI in multi‐site cohorts. Magnetic Resonance in Medicine. 2025;94(6):2732-2744. doi:10.1002/mrm.70020

BibTeX
@article{Saca2025,
  author = {Saca, Lucas and Gaggar, Raghav and Pappas, Ioannis and Benzinger, Tammie and Reiman, Eric M and Shiroishi, Mark S and Joe, Elizabeth B and Ringman, John M and Yassine, Hussein N and Schneider, Lon S and Chui, Helena C and Nation, Daniel A and Zlokovic, Berislav V and Toga, Arthur W and Chakhoyan, Ararat and Barnes, Samuel},
  title = {Automatic detection of arterial input function for brain DCE ‐ MRI in multi‐site cohorts},
  journal = {Magnetic Resonance in Medicine},
  year = {2025},
  volume = {94},
  number = {6},
  pages = {2732--2744},
  doi = {10.1002/mrm.70020}
}
Kostenlos starten →