Arterial input function estimation compensating for inflow and partial voluming in dynamic contrast-enhanced MRI

C. Tseng, M.A. Nagtegaal, Matthias J P van Osch, Jaap Jaspers, Alejandra Méndez Romero, Piotr Wielopolski, M. Smits, F.M. Vos

Research output: Contribution to journalArticleScientificpeer-review

26 Downloads (Pure)

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.
Original languageEnglish
Article numbere5225
Number of pages14
JournalNMR in Biomedicine
Volume37
Issue number12
DOIs
Publication statusPublished - 2024

Keywords

  • arterial input function
  • DCE MRI
  • inflow effect
  • partial volume effect

Fingerprint

Dive into the research topics of 'Arterial input function estimation compensating for inflow and partial voluming in dynamic contrast-enhanced MRI'. Together they form a unique fingerprint.

Cite this