1338
A Non-invasive Method for Quantifying Cerebral Blood Flow by Hybrid PET/MR1Lawson Health Research Institute, London, ON, Canada, 2Medical Biophysics, Western University, London, ON, Canada
Synopsis
While PET with [15O]H2O is the gold standard for imaging CBF, quantification requires measuring the arterial input function (AIF), which is an invasive and noisy procedure. ASL is an attractive alternative, however, its accuracy is limited by low SNR. Considering these limitations, we propose a hybrid PET/MRI approach using global CBF measurements from phase contrast MRI to convert [15O]H2O PET data into CBF maps. To test this method, using a large animal model, CBF was measured by this hybrid approach and by PET only, where the AIF was measured. Good agreement was found over a CBF range (20-100 ml/100g/min).
Introduction
Positron emission tomography (PET) using the tracer [15O]H2O is considered the gold standard for imaging cerebral blood flow (CBF). However, quantification requires measuring the arterial input function (AIF), which is not only invasive, but an inherently noisy procedure. Arterial spin labeling (ASL) is an attractive MRI-based alternative, but its accuracy is hindered by low signal to noise and arrival time uncertainties1,2, particularly when imaging CBF in patients with cerebrovascular disease. In light of these limitations, we propose a hybrid PET/MR approach that does not require invasive arterial sampling, but still generates quantitative CBF images. With this approach, global CBF is measured by phase-contrast MRI (PC-MRI)3 simultaneously with PET imaging of [15O]H2O. Global CBF is then used as a reference to convert [15O]H2O data into CBF maps, thereby avoiding the need to measure the AIF. This is similar to a proposed PET-only technique4, but with the important difference that global CBF is measured rather than assumed. In this study, the agreement between simultaneously measured CBF using PC-MRI and PET-only were compared in a large animal model over a range of CBF values.Methods
Experiments were approved by the Western University Research Ethics Board committee. Data were acquired in juvenile pigs (21.7 ± 2. kg) at baseline (n = 4), hypercapnia (n = 3) and hypocapnia (n = 3). Animals were anesthetized and PCO2 was manipulated by adjusting the breathing rate and volume. After a rapid intravenous bolus injection of [15O]H2O (475 ± 153 MBq), 5 min of PET list-mode data were acquired with a 3T hybrid PET/MR system (Biograph mMR Siemens). Arterial blood was sampled at 5mL/min using an MR-compatible blood sampling system (Swisstrace, Switzerland). For PC-MRI, an imaging plane orthogonal to the internal carotid and basilar arteries was identified by time-of-flight angiography (TR/TE: 22/3.6ms, matrix: 320 x 320 x 105, voxel size: 0.8 x 0.8 x 1.5 mm3). ECG-gated PC images were acquired during PET scanning (TR/TE: 34.4/2.87ms, matrix: 320 x 320, voxel size: 0.625 x 0.625 x 5 mm3, VENC: 80 cm/s in the through-plane direction, 8 averages). For structural reference, sagittal MPRAGE T1-weighted images were acquired (TR/TE: 1780/2.45ms, matrix: 256 x 256 x 176, voxel size: 1mm isotropic). Average whole brain flow was measured with PC-MRI data by contouring the feeding arteries in Argus Flow and scaling by the brain tissue weight. Raw PET data were reconstructed into 37 dynamic frames (3s x 20; 5s x 6; 15s x 6; 30s x 5) using a CT-based attenuation correction map and an ordered subset expectation maximization algorithm (matrix size: 344 x 344 x 127, voxel-size: 0.84 x 0.84 x 2 mm3). The images were smoothed with a 6-mm Gaussian filter, and a non-linear optimization routine was used to fit the Kety model5 (including the blood volume term) and determine CBF using a blood-brain partition coefficient of water = 90 ml/100g.Results
Mean pCO2 at hypocapnic, normocapnic and hypercapnic conditions were 25.0 ± 2.2, 37.6 ± 1.7 and 54.3 ± 4.9 mmHg. Average whole-brain CBF values measured by PC-MRI and O15-PET are summarized in Table 1. Representative CBF images at the 3 pCO2 conditions from the MRI-reference PET method are shown in Figure 1. A Linear regression and Bland-Altman plots are shown in Figures 2 and 3 to assess the agreement between the PET-based and PC-MRI measurements of CBF.Discussion and Conclusion
This
work presents a non-invasive and quantitative method of imaging CBF
by hybrid PET/MR. We believe this method could be useful for patient
populations for whom it has proven challenging to obtain accurate
perfusion measurements with other methods, most notably ASL, due to
significant vascular disease. CBF maps shown in Figure 1 demonstrate
the expected global perfusion increase and decrease in the hyper- and
hypocapnic states. Linear regression showed significant positive
correlation (R2=0.92, p < 0.05). The Bland-Altman plot
demonstrated good agreement between PET and MRI measurements (NS,
p = 0.22). Interestingly, in the hypercapnic state, our preliminary
data shows greater deviation between CBF measurements from PC-MRI and
PET (13%). This may be attributed to limited water extraction at
higher flow rates6. Future
studies will involve comparing this reference method to ASL in CVD
patients to assess its ability to quantify perfusion abnormalities.Acknowledgements
No acknowledgement found.References
1. Heijtel, D. F. R. et al. Accuracy and precision of pseudo-continuous arterial spin labeling perfusion during baseline and hypercapnia: a head-to-head comparison with 15O H2O positron emission tomography. Neuroimage 92, 182–92 (2014).
2. Haga, S. et al. Arterial Spin Labeling Perfusion Magnetic Resonance Image with Dual Postlabeling Delay: A Correlative Study with Acetazolamide Loading 123I-Iodoamphetamine Single-Photon Emission Computed Tomography. J. Stroke Cerebrovasc. Dis. 25, 1–6 (2015).
3. Peng, S. L. et al. Optimization of phase-contrast MRI for the quantification of whole-brain cerebral blood flow. J. Magn. Reson. Imaging 1–8 (2015).
4. Mejia, M. A. et al. Simplified nonlinearity correction of oxygen-15-water regional cerebral blood flow images without blood sampling. J. Nucl. Med. 35, 1870–7 (1994).
5. Kety, S. S. The theory and applications of the exchange of inert gas at the lungs and tissues. Pharmacol. Rev. 3, 1–41 (1951).
6. St Lawrence, K. S. et al. An adiabatic approximation to the tissue homogeneity model for water exchange in the brain: II. Experimental validation. J. Cereb. Blood Flow Metab. 18, 1378–1385 (1998).
Figures
