View in EDS HTML Full Text PDF Full Text

A physics‐informed deep learning framework for dynamic susceptibility contrast perfusion MRI.

Saved in:
Bibliographic Details
Title: A physics‐informed deep learning framework for dynamic susceptibility contrast perfusion MRI.
Authors: Rotkopf, Lukas T.1 (AUTHOR), Ziener, Christian H.1 (AUTHOR), von Knebel‐Doeberitz, Nikolaus1 (AUTHOR), Wolf, Sabine D.2 (AUTHOR), Hohmann, Anja3 (AUTHOR), Wick, Wolfgang3 (AUTHOR), Bendszus, Martin4 (AUTHOR), Schlemmer, Heinz‐Peter1 (AUTHOR), Paech, Daniel1 (AUTHOR), Kurz, Felix T.1,4,5 (AUTHOR) felixtobias.kurz@hug.ch
Source: Medical Physics. Dec2024, Vol. 51 Issue 12, p9031-9040. 10p.
Subjects: Perfusion magnetic resonance imaging, Singular value decomposition, Cerebral circulation, Perfusion imaging, Contrast media, Deep learning
Abstract: Background: Perfusion magnetic resonance imaging (MRI)s plays a central role in the diagnosis and monitoring of neurovascular or neurooncological disease. However, conventional processing techniques are limited in their ability to capture relevant characteristics of the perfusion dynamics and suffer from a lack of standardization. Purpose: We propose a physics‐informed deep learning framework which is capable of analyzing dynamic susceptibility contrast perfusion MRI data and recovering the dynamic tissue response with high accuracy. Methods: The framework uses physics‐informed neural networks (PINNs) to learn the voxel‐wise TRF, which represents the dynamic response of the local vascular network to the contrast agent bolus. The network output is stabilized by total variation and elastic net regularization. Parameter maps of normalized cerebral blood flow (nCBF) and volume (nCBV) are then calculated from the predicted residue functions. The results are validated using extensive comparisons to values derived by conventional Tikhonov‐regularized singular value decomposition (TiSVD), in silico simulations and an in vivo dataset of perfusion MRI exams of patients with high‐grade gliomas. Results: The simulation results demonstrate that PINN‐derived residue functions show a high concordance with the true functions and that the calculated values of nCBF and nCBV converge towards the true values for higher contrast‐to‐noise ratios. In the in vivo dataset, we find high correlations between conventionally derived and PINN‐predicted perfusion parameters (Pearson's rho for nCBF: 0.84±0.03$0.84 \pm 0.03$ and nCBV: 0.92±0.03$0.92 \pm 0.03$) and very high indices of image similarity (structural similarity index for nCBF: 0.91±0.03$0.91 \pm 0.03$ and for nCBV: 0.98±0.00$0.98 \pm 0.00$). Conclusions: PINNs can be used to analyze perfusion MRI data and stably recover the response functions of the local vasculature with high accuracy. [ABSTRACT FROM AUTHOR]
Copyright of Medical Physics is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: egs
DbLabel: Engineering Source
An: 181778089
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: A physics‐informed deep learning framework for dynamic susceptibility contrast perfusion MRI.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Rotkopf%2C+Lukas+T%2E%22">Rotkopf, Lukas T.</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Ziener%2C+Christian+H%2E%22">Ziener, Christian H.</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22von+Knebel‐Doeberitz%2C+Nikolaus%22">von Knebel‐Doeberitz, Nikolaus</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wolf%2C+Sabine+D%2E%22">Wolf, Sabine D.</searchLink><relatesTo>2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Hohmann%2C+Anja%22">Hohmann, Anja</searchLink><relatesTo>3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wick%2C+Wolfgang%22">Wick, Wolfgang</searchLink><relatesTo>3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Bendszus%2C+Martin%22">Bendszus, Martin</searchLink><relatesTo>4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Schlemmer%2C+Heinz‐Peter%22">Schlemmer, Heinz‐Peter</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Paech%2C+Daniel%22">Paech, Daniel</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Kurz%2C+Felix+T%2E%22">Kurz, Felix T.</searchLink><relatesTo>1,4,5</relatesTo> (AUTHOR)<i> felixtobias.kurz@hug.ch</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Medical+Physics%22">Medical Physics</searchLink>. Dec2024, Vol. 51 Issue 12, p9031-9040. 10p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Perfusion+magnetic+resonance+imaging%22">Perfusion magnetic resonance imaging</searchLink><br /><searchLink fieldCode="DE" term="%22Singular+value+decomposition%22">Singular value decomposition</searchLink><br /><searchLink fieldCode="DE" term="%22Cerebral+circulation%22">Cerebral circulation</searchLink><br /><searchLink fieldCode="DE" term="%22Perfusion+imaging%22">Perfusion imaging</searchLink><br /><searchLink fieldCode="DE" term="%22Contrast+media%22">Contrast media</searchLink><br /><searchLink fieldCode="DE" term="%22Deep+learning%22">Deep learning</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Background: Perfusion magnetic resonance imaging (MRI)s plays a central role in the diagnosis and monitoring of neurovascular or neurooncological disease. However, conventional processing techniques are limited in their ability to capture relevant characteristics of the perfusion dynamics and suffer from a lack of standardization. Purpose: We propose a physics‐informed deep learning framework which is capable of analyzing dynamic susceptibility contrast perfusion MRI data and recovering the dynamic tissue response with high accuracy. Methods: The framework uses physics‐informed neural networks (PINNs) to learn the voxel‐wise TRF, which represents the dynamic response of the local vascular network to the contrast agent bolus. The network output is stabilized by total variation and elastic net regularization. Parameter maps of normalized cerebral blood flow (nCBF) and volume (nCBV) are then calculated from the predicted residue functions. The results are validated using extensive comparisons to values derived by conventional Tikhonov‐regularized singular value decomposition (TiSVD), in silico simulations and an in vivo dataset of perfusion MRI exams of patients with high‐grade gliomas. Results: The simulation results demonstrate that PINN‐derived residue functions show a high concordance with the true functions and that the calculated values of nCBF and nCBV converge towards the true values for higher contrast‐to‐noise ratios. In the in vivo dataset, we find high correlations between conventionally derived and PINN‐predicted perfusion parameters (Pearson's rho for nCBF: 0.84±0.03$0.84 \pm 0.03$ and nCBV: 0.92±0.03$0.92 \pm 0.03$) and very high indices of image similarity (structural similarity index for nCBF: 0.91±0.03$0.91 \pm 0.03$ and for nCBV: 0.98±0.00$0.98 \pm 0.00$). Conclusions: PINNs can be used to analyze perfusion MRI data and stably recover the response functions of the local vasculature with high accuracy. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Medical Physics is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=181778089
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1002/mp.17415
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 10
        StartPage: 9031
    Subjects:
      – SubjectFull: Perfusion magnetic resonance imaging
        Type: general
      – SubjectFull: Singular value decomposition
        Type: general
      – SubjectFull: Cerebral circulation
        Type: general
      – SubjectFull: Perfusion imaging
        Type: general
      – SubjectFull: Contrast media
        Type: general
      – SubjectFull: Deep learning
        Type: general
    Titles:
      – TitleFull: A physics‐informed deep learning framework for dynamic susceptibility contrast perfusion MRI.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Rotkopf, Lukas T.
      – PersonEntity:
          Name:
            NameFull: Ziener, Christian H.
      – PersonEntity:
          Name:
            NameFull: von Knebel‐Doeberitz, Nikolaus
      – PersonEntity:
          Name:
            NameFull: Wolf, Sabine D.
      – PersonEntity:
          Name:
            NameFull: Hohmann, Anja
      – PersonEntity:
          Name:
            NameFull: Wick, Wolfgang
      – PersonEntity:
          Name:
            NameFull: Bendszus, Martin
      – PersonEntity:
          Name:
            NameFull: Schlemmer, Heinz‐Peter
      – PersonEntity:
          Name:
            NameFull: Paech, Daniel
      – PersonEntity:
          Name:
            NameFull: Kurz, Felix T.
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 12
              Text: Dec2024
              Type: published
              Y: 2024
          Identifiers:
            – Type: issn-print
              Value: 00942405
          Numbering:
            – Type: volume
              Value: 51
            – Type: issue
              Value: 12
          Titles:
            – TitleFull: Medical Physics
              Type: main
ResultId 1