0961

Dynamic resting state fMRI in mice: detection of Quasi-Periodic Patterns
Michaël Belloy1, Maarten Naeyaert1, Georgios Keliris1, Anzar Abbas2, Shella Keilholz2, Annemie Van Der Linden1, and Marleen Verhoye1

1Bio-Imaging Lab, Antwerpen, Belgium, 2MIND lab, Atlanta, GA

Synopsis

We report, to our knowledge, the first application of a dynamic rsfMRI approach in mice and show the reliable detection of Quasi-Periodic Patterns (QPP) at both the group-level as well as in single-subject data. These patterns are consistent with QPPs detected previously in humans and rats, displaying a high intensity wave spreading across the cortex from lateral towards upward medial, followed by a low intensity wave in the same direction. These findings are promising for the application of QPPs towards investigating mouse resting state functional connectivity and its development as a potential new pre-clinical tool.

Introduction

Recently, a vast body of emerging literature has indicated that dynamic over ‘static’ analysis of temporal features in BOLD resting state (rs) functional connectivity, as measured with fMRI, provides many additional insights into the macroscale organization and dynamics of neural activity1,2. This drives both fundamental research into studying features of large-scale brain functioning, while also enabling new (pre-)clinical applications for the investigation and detection of neurological disorders. These studies however remained mostly limited to humans, and in lesser extent to monkeys and rats3,4. No work in mice has been reported, yet the high throughput and wide range of available mouse disease models compels for exploration. Here, we present, to our knowledge, the first application of a dynamic rsfMRI approach to mouse resting state data and show the reliable detection of Quasi-Periodic Patterns (QPP) in the BOLD fluctuations, both on a group and single subject level.

Methods

RsfMRI was acquired in 11 five-month old C57BL/6 mice, on a 9.4T Biospec scanner, using a high temporal resolution EPI sequence (repetitions:1200; TR:500ms; TE:16ms; FA:55°; FOV:(20x20)mm2; matrix:128x64; 3 axial slices; thickness/separation:0.4mm/0.1mm). Slices were positioned centered around the somatosensory cortex. Animals were anesthetized with isoflurane (0.4%) and medetomidine bolus (0.1mg/kg) + continuous infusion (0.05mg/kg), starting at 15min post-bolus. Scans were acquired 40min post-bolus.

For group-level analysis, data was smoothed, realigned and normalized to a user-specified template animal, using Statistical Paramagnetic Mapping 12 (SPM12), followed by band pass filtering (0.01-0.2Hz), quadratic detrending, demeaning, and application of a ventricular mask. The preprocessed middle slice image series of all animals were then concatenated, before passing them to the QPP detection algorithm5 (see Fig.1). For single subject analysis, no normalization was performed and the image series were individually passed to the algorithm. All analysis was performed in MATLAB2015a (Mathworks).

Results

The QPP detection algorithm at the group-level was run at different window sizes and was repeated 500 times for each, followed by kmeans clustering and silhouette detection to determine the respective ideal QPPs. All QPPs showed a high spread of occurrences throughout the different subjects (see Fig.2 for illustration), while an increase in window size correlated with a decrease in QPP frequency.

Inspection of all obtained QPPs indicated the presence of a common pattern, which was however displayed at different subsections and offsets, depending on the window size. The different representations of QPPs can be better understood from the offsets in their correlation time series (see Fig.3), illustrating how the different QPPs are detected at other starting time points.

At a window size of 27 images, we inferred a full non-redundant coverage with a representative starting time point (see Fig.4). The pattern shows a high intensity wave from lateral towards upward medial, followed by a low intensity wave in the same direction. The other QPP, due to their quasi-periodic nature, can be circularly shifted to match the start of QPP27 (see Fig.4).

Further, single subject analysis with the same window sizes showed QPP patterns and detection time points consistent with those observed in the same subject, analyzed at the group level (see Fig.5).

Discussion & Conclusion

We present an initial framework to interpret the detection of Quasi-Periodic Patterns (QPP) in mouse rsfMRI data. With this, we show the reliable detection of QPPs at both the group-level as well as in single-subject data. The latter is extremely interesting towards the development of more reliable rsfMRI techniques and analysis strategies in (pre-clinical) mice studies. Furthermore, the observed patterns are consistent with QPPs previously detected in humans and rats, where they were linked with infraslow local field potentials, indicating preservation between species and a relationship with an underlying neural basis5,6. These findings are promising for the application of QPPs towards investigating mouse resting state functional connectivity and its development as a potential pre-clinical biomarker.

Work is currently ongoing to fully understand these QPPs and how they relate to the functional connectivity observed in rsfMRI, both in control animals and in mouse disease models. Additionally, we are investigating the potential confounds induced by physiological parameters.

Acknowledgements

No acknowledgement found.

References

1. Deco G, Jirsa VK, Mcintosh AR. Emerging concepts for the dynamical organization of resting-state activity in the brain. 2011;12(january). doi:10.1038/nrn2961.

2. Hutchison RM, Womelsdorf T, Allen E a., et al. Dynamic functional connectivity: Promise, issues, and interpretations. Neuroimage. 2013;80:360-378. doi:10.1016/j.neuroimage.2013.05.079.

3. Keilholz SD. The neural basis of time-varying resting-state functional connectivity. Brain Connect. 2014;4(10):769-779. doi:10.1089/brain.2014.0250.

4. Keilholz SD, Magnuson ME, Pan W-J, Willis M, Thompson GJ. Dynamic properties of functional connectivity in the rodent. Brain Connect. 2013;3(1):31-40. doi:10.1089/brain.2012.0115.

5. Majeed W, Magnuson M, Hasenkamp W, et al. Spatiotemporal dynamics of low frequency BOLD fluctuations in rats and humans. Neuroimage. 2011;54(2):1140-1150. doi:10.1016/j.neuroimage.2010.08.030.Spatiotemporal.

6. Pan WJ, Thompson GJ, Magnuson ME, Jaeger D, Keilholz S. Infraslow LFP correlates to resting-state fMRI BOLD signals. Neuroimage. 2013;74:288-297. doi:10.1016/j.neuroimage.2013.02.035.

Figures

Fig.1 Visual representation of the Quasi-Periodic Pattern (QPP) detection algorithm5 (adapted from Majeed et al., 2011). Briefly, a randomly selected image sequence with a specified window length is slid over the preprocessed image series and compared to create a correlation time series. Image sequences detected at threshold crossings are then averaged into an updated template. This process is repeated again by sliding the updated template over the image series, further updating the template, until it remains stable after x iterations. This creates the QPP and the entire process can be repeated to allow for different random starting time points.

Fig.2 Anatomical RARE image (top left) of the center slice under study. The Quasi-Periodic Pattern (QPP, upper panel) and the sliding window correlation time series of this QPP with the concatenated preprocessed image series of all subjects (lower panel), observed for a window size of 18 images. The QPP, which is displayed with a 3-images or 1.5s intersperse, indicates a high intensity wave from lateral towards upward medial, followed by a low intensity wave in the same direction. All peaks crossing above a 0.2 correlation threshold (triangles) indicate occurences of the QPP.

Fig.3 Illustration of the overlap between group-level Quasi-Period Patterns (QPP) sliding window correlation time series observed at different window sizes. Note the shifted correlation time traces in blue and the opposite phased ones in red.

Fig.4 Group-level Quasi-Periodic Patterns (QPP) as observed for all analyzed window sizes. The QPPs are circularly shifted to align with QPP27 (green box). The shifts are determined by the phase offsets, indicated on the crosscorrelograms (left panel) of the respective QPP correlation time series (see Fig.3) with the correlation time series of QPP27. Note the loss of observable patterns after 27 windows (red box).

Fig.5 A) The Quasi-Periodic Pattern (QPP) detected with window size 27 in a single-subject analysis. Note the similarity with the QPP detected at the same window size on the group-level (see fig.4). B) The QPP correlation time series observed with window size 27, for single subject (red) and same-subject within the group-level analysis (blue). Note the high overlap of peaks and greater stringency of the group level analysis leading to smaller correlation peaks.

Proc. Intl. Soc. Mag. Reson. Med. 25 (2017)
0961