Your browser doesn't support javascript.
loading
Arterial input function estimation compensating for inflow and partial voluming in dynamic contrast-enhanced MRI.
Tseng, Chih-Hsien; Nagtegaal, Martijn A; van Osch, Matthias J P; Jaspers, Jaap; Mendez Romero, Alejandra; Wielopolski, Piotr; Smits, Marion; Vos, Frans M.
Affiliation
  • Tseng CH; Department of Imaging Physics, Delft University of Technology, Delft, the Netherlands.
  • Nagtegaal MA; Medical Delta, Delft, the Netherlands.
  • van Osch MJP; HollandPTC Consortium-Erasmus MC, Rotterdam, Holland Proton Therapy Center, Delft, Leiden University Medical Center, Leiden and Delft University of Technology, Delft, the Netherlands.
  • Jaspers J; Department of Imaging Physics, Delft University of Technology, Delft, the Netherlands.
  • Mendez Romero A; C.J. Gorter MRI Center, Department of Radiology, Leiden University Medical Center, Leiden, the Netherlands.
  • Wielopolski P; Medical Delta, Delft, the Netherlands.
  • Smits M; HollandPTC Consortium-Erasmus MC, Rotterdam, Holland Proton Therapy Center, Delft, Leiden University Medical Center, Leiden and Delft University of Technology, Delft, the Netherlands.
  • Vos FM; C.J. Gorter MRI Center, Department of Radiology, Leiden University Medical Center, Leiden, the Netherlands.
NMR Biomed ; : e5225, 2024 Aug 06.
Article in En | MEDLINE | ID: mdl-39107878
ABSTRACT
Both inflow and the partial volume effect (PVE) are sources of error when measuring the arterial input function (AIF) in dynamic contrast-enhanced (DCE) MRI. This is relevant, as errors in the AIF can propagate into pharmacokinetic parameter estimations from the DCE data. A method was introduced for flow correction by estimating and compensating the number of the perceived pulse of spins during inflow. We hypothesized that the PVE has an impact on concentration-time curves similar to inflow. Therefore, we aimed to study the efficiency of this method to compensate for both effects simultaneously. We first simulated an AIF with different levels of inflow and PVE contamination. The peak, full width at half-maximum (FWHM), and area under curve (AUC) of the reconstructed AIFs were compared with the true (simulated) AIF. In clinical data, the PVE was included in AIFs artificially by averaging the signal in voxels surrounding a manually selected point in an artery. Subsequently, the artificial partial volume AIFs were corrected and compared with the AIF from the selected point. Additionally, corrected AIFs from the internal carotid artery (ICA), the middle cerebral artery (MCA), and the venous output function (VOF) estimated from the superior sagittal sinus (SSS) were compared. As such, we aimed to investigate the effectiveness of the correction method with different levels of inflow and PVE in clinical data. The simulation data demonstrated that the corrected AIFs had only marginal bias in peak value, FWHM, and AUC. Also, the algorithm yielded highly correlated reconstructed curves over increasingly larger neighbourhoods surrounding selected arterial points in clinical data. Furthermore, AIFs measured from the ICA and MCA produced similar peak height and FWHM, whereas a significantly larger peak and lower FWHM was found compared with the VOF. Our findings indicate that the proposed method has high potential to compensate for PVE and inflow simultaneously. The corrected AIFs could thereby provide a stable input source for DCE analysis.
Key words

Full text: 1 Collection: 01-internacional Database: MEDLINE Language: En Journal: NMR Biomed Journal subject: DIAGNOSTICO POR IMAGEM / MEDICINA NUCLEAR Year: 2024 Type: Article Affiliation country: Netherlands

Full text: 1 Collection: 01-internacional Database: MEDLINE Language: En Journal: NMR Biomed Journal subject: DIAGNOSTICO POR IMAGEM / MEDICINA NUCLEAR Year: 2024 Type: Article Affiliation country: Netherlands