Dennis Hedderich1, Anne Kluge1, Thomas Pyka2, Claus Zimmer1, Jan S Kirschke1, Benedikt Wiestler1, Christine Preibisch3. 1. Department of Diagnostic and Interventional Neuroradiology, Technische Universität München, Ismaningerst. 22, 81675 Munich, Germany. 2. Clinic for Nuclear Medicine, Technische Universität München, Ismaningerst. 22, 81675 Munich, Germany. 3. Department of Diagnostic and Interventional Neuroradiology, Technische Universität München, Ismaningerst. 22, 81675 Munich, Germany; Clinic for Neurology, Technische Universität München, Ismaningerst. 22, 81675 Munich, Germany. Electronic address: preibisch@tum.de.
Abstract
BACKGROUND AND PURPOSE: Several leakage correction algorithms for dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI)-based cerebral blood volume (CBV) measurement have been proposed, and combination with a preload of contrast agent is generally recommended. A single bolus application scheme would largely simplify and facilitate standardized clinical applications, while reducing contrast agent (CA) dose. The aim of this study was, therefore, to investigate whether appropriate leakage correction redundantizes prebolus application by comparing normalized DSC-based CBV (nCBV) measures of two consecutive CA boli. MATERIALS AND METHODS: Twenty-seven patients with suspected glioblastoma (WHO-grade-IV) underwent DSC-MRI during two consecutive boli of Gd-based CA. Four variants of two post-processing leakage correction techniques were compared with respect to nCBV in contrast enhancing tumor tissue. First, a reference curve approach with first pass and full integration of corrected ΔR2*(t), and second, a deconvolution-based approach using singular value decomposition (SVD) with a standard noise-dependent cutoff or Tikhonov regularization. RESULTS: Compared to respective uncorrected values, all leakage correction techniques increased nCBV for data acquired without prebolus, while there was no consistent trend for data acquired with prebolus. The best agreement between corrected nCBV values in contrast enhancing tumor, obtained in the same patients without and with prebolus, respectively, was obtained for the reference curve-based correction approach with either first pass or full integration. CONCLUSION: The reference curve-based leakage correction approach with integration-based nCBV calculation yielded a high accordance between nCBV values without and with prebolus, respectively. Thus, it appears possible to obtain valid nCBV in glioblastoma with a single CA injection.
BACKGROUND AND PURPOSE: Several leakage correction algorithms for dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI)-based cerebral blood volume (CBV) measurement have been proposed, and combination with a preload of contrast agent is generally recommended. A single bolus application scheme would largely simplify and facilitate standardized clinical applications, while reducing contrast agent (CA) dose. The aim of this study was, therefore, to investigate whether appropriate leakage correction redundantizes prebolus application by comparing normalized DSC-based CBV (nCBV) measures of two consecutive CA boli. MATERIALS AND METHODS: Twenty-seven patients with suspected glioblastoma (WHO-grade-IV) underwent DSC-MRI during two consecutive boli of Gd-based CA. Four variants of two post-processing leakage correction techniques were compared with respect to nCBV in contrast enhancing tumor tissue. First, a reference curve approach with first pass and full integration of corrected ΔR2*(t), and second, a deconvolution-based approach using singular value decomposition (SVD) with a standard noise-dependent cutoff or Tikhonov regularization. RESULTS: Compared to respective uncorrected values, all leakage correction techniques increased nCBV for data acquired without prebolus, while there was no consistent trend for data acquired with prebolus. The best agreement between corrected nCBV values in contrast enhancing tumor, obtained in the same patients without and with prebolus, respectively, was obtained for the reference curve-based correction approach with either first pass or full integration. CONCLUSION: The reference curve-based leakage correction approach with integration-based nCBV calculation yielded a high accordance between nCBV values without and with prebolus, respectively. Thus, it appears possible to obtain valid nCBV in glioblastoma with a single CA injection.
Authors: Shivaram Avula; Andrew Peet; Giovanni Morana; Paul Morgan; Monika Warmuth-Metz; Tim Jaspan Journal: Childs Nerv Syst Date: 2021-05-10 Impact factor: 1.475
Authors: S Schön; J Cabello; F Liesche-Starnecker; M Molina-Romero; P Eichinger; M Metz; I Karimov; C Preibisch; J Keupp; A Hock; B Meyer; W Weber; C Zimmer; T Pyka; I Yakushev; J Gempt; B Wiestler Journal: Eur J Nucl Med Mol Imaging Date: 2020-01-17 Impact factor: 9.236
Authors: K J Paprottka; S Kleiner; C Preibisch; F Kofler; F Schmidt-Graf; C Delbridge; D Bernhardt; S E Combs; J Gempt; B Meyer; C Zimmer; B H Menze; I Yakushev; J S Kirschke; B Wiestler Journal: Eur J Nucl Med Mol Imaging Date: 2021-06-25 Impact factor: 9.236