Calculation of cerebral perfusion parameters using regional arterial input functions identified by factor analysis

Area of Science:

• Neuroimaging
• Medical Physics
• Radiology

Background:

• Accurate calculation of regional cerebral blood volume (rCBV), regional cerebral blood flow (rCBF), and regional mean transit time (rMTT) requires a precise arterial input function (AIF).
• Traditional methods often face challenges with dispersion and delay when using a single AIF for the entire brain.

Purpose of the Study:

• To develop and validate a novel method for identifying multiple arterial input functions (AIFs) using Factor Analysis of Dynamic Studies (FADS).
• To improve the accuracy of pixel-by-pixel cerebral perfusion calculations by utilizing the nearest AIF to each voxel.

Main Methods:

• Simulated dynamic contrast-enhanced (DCE) MRI data were generated with added arterial and tissue dispersion/delay.
• Factor Analysis of Dynamic Studies (FADS) was employed to identify multiple AIFs.
• Cerebral perfusion parameters were calculated pixel-by-pixel using the nearest identified AIF.
• Simulations were performed across various signal-to-noise ratios (SNRs) and compared with single-pixel manual AIF selection.
• In vivo studies were conducted on healthy volunteers and patients.

Main Results:

• The FADS method effectively identified multiple AIFs, mitigating underestimation of rCBF caused by dispersion or delay.
• The nearest-AIF approach demonstrated robustness across different SNRs and simulated physiological variations.
• In vivo results corroborated the improved accuracy and reliability of the FADS-based method.

Conclusions:

• The integration of FADS with the nearest-AIF selection strategy offers a significant advancement over manual selection of a single AIF.
• This method enhances the accuracy of quantitative cerebral perfusion imaging, particularly in the presence of physiological complexities.