Proceedings of The Physiological Society

Physiology 2015 (Cardiff, UK) (2015) Proc Physiol Soc 34, C69

Oral Communications

Mapping the visual pathway in the mouse brain using snapshot fMRI

A. Niranjan1, J. A. Wells1, M. Lythgoe1

1. Centre for Advanced Biomedical Imaging, University College London, London, England, United Kingdom.


  • Fixed effects analysis (two-tailed t-test, p <0.05, FWE corrected) for a single subject, overlaid on raw GE-EPI of the same slice taken with multiple snapshots (1-4), with FSEMS anatomical reference scan above for comparison.

Introduction: Mouse brain fMRI is difficult due to the requirement for small voxels, and image distortion due to B0 inhomogeneities [1, 2]. The visual pathway represents an important research target for mapping in clinical and basic science applications. Previous task-based studies in the mouse have focused on electrical stimulation of the fore/hind paw [3]. However, visual stimulation in mouse fMRI is unexplored, with the only existing data [4] reporting atypical BOLD responses when compared to rats and humans. In this work, we characterise the BOLD response to visual stimulation in the mouse brain. Furthermore, we also explore the benefits of snapshot GE-EPI to reduce image distortion. This represents the first application of snapshot EPI to mouse brain task-based fMRI. Methods: Female C57BL6/J mice ((19.0 ± 0.6) g, N = 6) were imaged under medetomidine anaesthesia (0.4 mg/kg for initial bolus, 0.8 mg/kg/hr for constant infusion). A 9.4T MRI scanner (Agilent Inc.) was used with a volume coil for RF transmission (Rapid), and a 2-channel array mouse head surface coil (Rapid) for signal reception. fMRI data were acquired using GE-EPI with 1 to 4 snapshots. For visual stimulation, a fibre optic cable was placed in between the mouse head and the surface coil. The stimulus consisted of 445 nm light from a laser (Omicron) scattered off the inside of the surface coil. The fMRI paradigm consisted of 40 seconds of rest followed by 20 seconds of visual stimulation (10 Hz, 10 ms pulse duration, 0.3 mW (time averaged)), repeated three times per scan. Acquisition of fMRI data was conducted twice at each number of snapshots in an interleaved, pseudorandom manner. Data analysis was completed using the SPM12 MATLAB toolbox. Results: fMRI activation maps overlaid on GE-EPI images for a single animal are shown in Fig. 1. As the number of shots used in the GE-EPI increases, the distortion due to B0 inhomogeneities is reduced. Using a positive BOLD contrast, activation was detected in the superior colliculus (SC) and lateral geniculate nuclei (LGN) in all subjects using fixed effects analysis (p < 0.05, FWE corrected). These results agree with similar experiments conducted in rats [5]. Conclusions: We demonstrate a robust fMRI response to visual stimulation in the mouse brain using a free breathing, recoverable protocol with medetomidine anaesthesia, suitable for longitudinal protocols.

Where applicable, experiments conform with Society ethical requirements