Neuroimaging Research

Time-resolved functional PET-MRI fusion: temporally coupled metabolic and BOLD dynamics across task and naturalistic arousal

Sean Courseya,b, Shirley Fengb, Jingyuan E. Chenb,c

a  College of Science, Northeastern University, Boston, MA, USA;

b Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, USA;

c Department of Radiology, Harvard Medical School, Boston, MA, USA;

Introduction

In the realm of functional imaging, the integration of PET-MR systems and the introduction of continuous infusion FDG-PET (fPET-FDG) has recently permitted the concurrent examination of the dynamics of glucose metabolism and blood-oxygenation-level-dependent (BOLD) signals within the human brain[1-3]. Despite the most revolutionary aspect of fPET being its high temporal resolution comparative to previous PET methods[4], most fPET-fMRI studies to date have centered on static spatial correlations rather than the dynamic interplay between these signals over time[5-7]. Capitalizing on fPET-FDG’s ability to capture dynamic changes in glucose uptakes, our study introduces an analytical framework that connects the temporal evolution of fPET-FDG signals with real-time BOLD-fMRI data. This methodology was tested against two datasets—one incorporating visual stimulation and another utilizing a naturalistic arousal paradigm—to determine its effectiveness in linking fPET-FDG dynamics with concurrent BOLD-fMRI time-courses.

Model construction & validation I — visual task dataset

We first assessed the temporal coupling between BOLD-fMRI and FDG-PET measures in a well-controlled task paradigm, by analyzing a publicly available fPET-fMRI dataset employing visual stimulation[8]. The Monash vis-fPET-fMRI dataset comprises 70-minute simultaneous FDG-fPET and BOLD-fMRI scans of 10 healthy young adults (fMRI: voxel size = 3x3x3 mm3, TR = 2.45 s; fPET: reconstructed nominal voxel size = 1.39x1.392 mm3, temporal resolution = 1 min). Using an embedded block design, participants were exposed to a visual stimulus consisting of a flickering checkerboard (Fig. 1). 

We hypothesize that the power of the fMRI signal at the stimulus frequency is indicative of neural activity, so integrating this signal over time will correspond to the concurrent FDG accumulation due to metabolic demand. To test this, we computed the average fMRI signal from the primary visual cortex (V1) across subjects, applied a Hilbert transform to determine the BOLD signal power over time, and integrated this power to use as a predictor variable for the FDG-PET data (Fig. 1). We removed long-term trends from both the predictor (regressor) and the PET data using a third-order polynomial, focusing on the dynamic changes. Subsequent regression analysis showed that our integrated BOLD signal predictor aligns well with the FDG accumulation in the detrended V1 area (Fig. 1, fMRI-modeled fPET dynamics vs. detrended fPET signal). Moreover, the significant correlation between the BOLD signal and FDG accumulation is specifically strong in the visual cortex, as shown in the spatial distribution of the test statistics (Fig. 2, highlighted by green arrows). We also noticed metabolic activations in the frontal cortex, possibly owing to task-locked attention changes.

Model construction & validation II — sleep imaging dataset

To further evaluate our model of BOLD-FDG temporal coupling for a naturalistic paradigm, we next acquired and analyzed fPET-fMRI data from 23 healthy adults who were instructed to close their eyes and relax throughout a 75-120 min scan. The subject’s arousal states were inferred from simultaneous EEG recordings (18 subjects) or behavioral measures (5 subjects). All scans were performed on a 3T Siemens MR scanner with a BrainPET insert (fMRI: voxel size = 3x3x3 mm3, TR = 2/2.4 s; fPET: reconstructed nominal voxel size = 2.5x2.5x2.5 mm3, temporal resolution = 30 s). FDG was administered with a bolus plus continuous infusion paradigm, with the bolus comprising 20% of the total continuously infused dose.

As shown in Fig. 3, our tri-modal framework could successfully track sleep-wake dynamics in both glucose uptakes and hemodynamic changes. Following a similar framework as the visual dataset, we identified a strong coupling between the global BOLD-fMRI and FDG time-course (Fig. 4). The metabolic regressor modeled by the global fMRI signal explained considerable variance of fPET time-activity curves in extensive cortical regions (Fig. 5). These observations supported the applicability of the PET-MRI integration framework in naturalistic paradigms.  

Conclusion

In this study, we proposed and validated a framework that links concurrent dynamics of metabolic and hemodynamic signals. Using this framework, concurrent fMRI signals could successfully predict fPET-FDG dynamics driven by both explicit visual stimuli and naturalistic arousal. Our results substantiate the dynamic coupling of metabolic and hemodynamic changes in the human brain, emphasizing the value of fPET-FDG in conjunction with BOLD-fMRI for characterizing time-resolved interdependence of neurovascular and neurometabolic activity. Consequently, our research advances the exploration of temporal couplings between hemodynamics and neural metabolism, addressing a theme of fundamental interest to neuroimaging and the wider field of neuroscience.

References 

[1] Villien, M., Wey, H.Y., Mandeville, J.B., Catana, C., Polimeni, J.R., Sander, C.Y., Zürcher, N.R., Chonde, D.B., Fowler, J.S., Rosen, B.R. and Hooker, J.M., 2014. Dynamic functional imaging of brain glucose utilization using fPET-FDG. Neuroimage, 100, pp.192-199.

[2] Hahn, A., Gryglewski, G., Nics, L., Hienert, M., Rischka, L., Vraka, C., Sigurdardottir, H., Vanicek, T., James, G.M., Seiger, R. and Kautzky, A., 2016. Quantification of task-specific glucose metabolism with constant infusion of 18F-FDG. Journal of Nuclear Medicine, 57(12), pp.1933-1940.

[3] Jamadar, S.D., Ward, P.G., Li, S., Sforazzini, F., Baran, J., Chen, Z. and Egan, G.F., 2019. Simultaneous task-based BOLD-fMRI and [18-F] FDG functional PET for measurement of neuronal metabolism in the human visual cortex. Neuroimage, 189, pp.258-266.

[4] Rischka, L., Gryglewski, G., Pfaff, S., Vanicek, T., Hienert, M., Klöbl, M., Hartenbach, M., Haug, A., Wadsak, W., Mitterhauser, M. and Hacker, M., 2018. Reduced task durations in functional PET imaging with [18F] FDG approaching that of functional MRI. Neuroimage, 181, pp.323-330.

[5] Stiernman, L.J., Grill, F., Hahn, A., Rischka, L., Lanzenberger, R., Panes Lundmark, V., Riklund, K., Axelsson, J. and Rieckmann, A., 2021. Dissociations between glucose metabolism and blood oxygenation in the human default mode network revealed by simultaneous PET-fMRI. Proceedings of the National Academy of Sciences, 118(27), p.e2021913118.

[6] Hahn, A., Breakspear, M., Rischka, L., Wadsak, W., Godbersen, G.M., Pichler, V., Michenthaler, P., Vanicek, T., Hacker, M., Kasper, S. and Lanzenberger, R., 2020. Reconfiguration of functional brain networks and metabolic cost converge during task performance. elife, 9, p.e52443.

[7] Jamadar, S.D., Ward, P.G., Liang, E.X., Orchard, E.R., Chen, Z. and Egan, G.F., 2021. Metabolic and hemodynamic resting-state connectivity of the human brain: a high-temporal resolution simultaneous BOLD-fMRI and FDG-fPET multimodality study. Cerebral Cortex, 31(6), pp.2855-2867.

[8] Jamadar, S.D., Zhong, S., Carey, A., McIntyre, R., Ward, P.G., Fornito, A., Premaratne, M., Jon Shah, N., O’Brien, K., Stäb, D. and Chen, Z., 2021. Task-evoked simultaneous FDG-PET and fMRI data for measurement of neural metabolism in the human visual cortex. Scientific Data, 8(1), p.267.

Figures

2023 OHBM Abstract

This abstract, which serves as an introduction to my research improving statistical methods for fPET-FDG analysis, was accepted for a poster presentation at the 2023 Organization for Human Brain Mapping Conference in Montreal.

Bias of metabolic (de)activations introduced by polynomial detrending of fPET-FDG data

Sean Courseya,b, Grant A. Hartungb,c, Jonathan R. Polimenib,c,d, Jingyuan E. Chenb,c

a  College of Science, Northeastern University, Boston, MA, USA;

b Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, USA;

c Department of Radiology, Harvard Medical School, Boston, MA, USA;

d Harvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, USA.

Introduction

Functional PET (fPET)-FDG is a new addition to the neuroimaging field that enables tracking dynamic changes of metabolic activity within a single scan (Hahn et al. 2016; Stiernman et al. 2021; Villien et al. 2014). Because FDG is administered through constant infusion and accumulates in cells over the course of the scan, the time activity curve (TAC) has a monotonically increasing baseline trend. To separate task-driven metabolic deviations from this baseline, previous studies have used various methods to remove the baseline, including detrending using the whole-brain mean TAC or polynomials. While polynomial detrending can help avoid removing meaningful global metabolic changes—an issue with using the mean TAC—polynomials can mischaracterize the shape of the TAC baseline such that the resulting regression leaves a consistently patterned residual time-course. Furthermore, when this residual time-course correlates with the task paradigm, it can introduce artifactual (de)activations that bias the statistical testing of the true task effect. Here, we combine simulations and analysis of empirical data to investigate such activation biases caused by baseline misspecification in fPET-FDG studies.

Methods

We first performed simulations to assess how the activation bias introduced by baseline misspecification varies with arterial input function (AIF), tissue biological parameters, the choice of baseline model, and task timing. Our simulation used a two-compartment model of the radiotracer kinetics (Villien et al. 2014). The biological ranges for the kinetic constants K1, k2, and k3 were taken from Heiss et al. 1984. The effect size of the activation bias was quantified for three different baseline models: whole-brain mean TAC and 2nd- and 3rd-order polynomials; and three different task paradigms associated with Regressors 1-3 shown in Fig. 1b. The fitting coefficient for the residual pattern was compared to that for the activation profile to calculate percent bias. Activation was modeled as a 30% increase in k3.

We used publicly available resting-state, i.e. task-free, fPET-FDG data from 25 subjects (Jamadar et al. 2020) to test whether mismodeled baseline trends can lead to spurious activations in real data. We applied the same three baseline models and task regressors from the simulations to this dataset and tested for significant metabolic activations at the group level.

Results

Using low-order polynomials to detrend resting-state data resulted in consistent residual patterns across subjects, and these residual patterns from the data agreed with simulations (Fig. 1a). Simulations indicated that the bias resulting from these patterns can be substantial but that the extent of this bias varies greatly depending on the task paradigm, polynomial order, kinetic constants, and AIF (Fig. 1b).

Fig. 2 shows the results of applying different task regressors to the resting-state data using mean-TAC and low-order polynomial baselines. The extent and spatial patterning of bias vary greatly, depending on the task paradigm and the polynomial order. For instance, Regressor 2 in conjunction with 2nd-order detrending produced significant artifactual activation in the right occipital lobe at the group level; and Regressor 3 in conjunction with 3rd-order detrending produced modest global artifactual activation. With a sufficiently large sample size, any bias introduced by baseline mischaracterization may potentially result in statistically significant activations at the group level (Smith et al. 2018). 

Conclusion

Detrending fPET-FDG data using polynomials can leave residual patterns which correlate with task-regressors, biasing analysis results and potentially leading to artifactual activation. This bias depends on the task paradigm, polynomial order, and tracer kinetics—ranging from insignificant to producing artifactual activation with various spatial patterns. These results underscore the importance of choosing detrending methods carefully in fPET-FDG studies.

Figures