Neuroscience Clinical Applications of fMRI

Wednesday 14 May 2014

Michelle Moerel¹, Federico De Martino², Roberta Santoro², Kamil Ugurbil¹, Essa Yacoub¹, and Elia Formisano^2
1Department of Radiology, Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, Minnesota, United States, ²Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, Netherlands

We combined computational modeling with ultra-high field fMRI (7T) to explore responses to natural sounds in two subcortical auditory structures: the inferior colliculus (IC) and the medial geniculate body (MGB). Out of four computational models tested, only a model with a dependent frequency by location tuning could significantly predict responses to novel testing sounds. This suggests that at the level of IC and MGB, neuronal populations jointly encode sound frequency and sound location. Color-coding voxels according to their best frequency and location revealed fine-grained tonotopic maps and tuning to contralateral sound locations.

Alexander Poplawsky¹, Mitsuhiro Fukuda¹, Xiaopeng Zong², and Seong-Gi Kim^1,3
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²Biomedical Research Imaging Center, University of North Carolina, Chapel Hill, NC, United States, ³Biological Sciences, Center for Neuroscience Imaging Research, Institute for Basic Science (IBS), SKKU, Suwon, Korea

We used high-resolution fMRI in the olfactory bulb model to examine whether hemodynamic responses are regulated at the laminar level. Odor and micro-stimulation of the lateral olfactory tract (LOT) were used to preferentially evoke activity in different bulb layers. We found different, neural-specific laminar activation patterns for each individual stimulation using CBV fMRI, but not BOLD. Specifically, for CBV, a ring-like activation was reproducibly observed only in middle layers (~200 µm thick) following LOT stimulation that corresponds with the location of evoked neural activity. Our data provide definitive evidence that the hemodynamic response is regulated at the laminar level.

Matthew C. Murphy^1,2, Christopher Fisher¹, Seong-Gi Kim^2,3, Joel S. Schuman¹, Amy C. Nau¹, and Kevin C. Chan^1,2
1Department of Ophthalmology, University of Pittsburgh, Pittsburgh, PA, United States, ²Neuroimaging Laboratory, Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ³Center for Neuroscience Imaging Research, Department of Biological Sciences, SKKU, Suwon, Korea

Blind persons may acquire visual perception indirectly using sensory substitution devices. One such device (The vOICe) converts images to auditory “soundscapes.” To examine the effect of top down input on the processing of soundscapes in the visual cortex, we scanned sighted and blind subjects both immediately before and immediately after training them to interpret the soundscapes as images. Significant negative BOLD responses were observed in the visual cortex of sighted subjects in both conditions. This negative response is absent in blind subjects before training. Furthermore, the BOLD response becomes significantly positive after training indicating of a significant top down influence.

Bistra Iordanova¹, Alberto T Vazquez¹, Alex Poplawsky¹, Takashi Kozai², Hiro Fukuda¹, Matthew Murphy¹, and Seong-Gi Kim^1,3
1Dept. of Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²Dept. of Bioengineering, University of Pittsburgh, Pittsburgh, PA, United States, ³Dept. of Biological Sciences, Center for Neuroscience Imaging Research, Suwon, Korea

We use optogenetic approach to modulate the activity of the pyramidal cells in the primary somatosensory cortex. We record hemoglobin-based optical intrinsic signal (OIS) at different light activation parameters. We then compare the OIS to blood oxygenated level dependent (BOLD) fMRI activation.

Andreas Bruns¹, Thomas Mueggler¹, Basil Künnecke¹, Céline Risterucci¹, Joseph G. Wettstein¹, and Markus von Kienlin^1
1Neuroscience Discovery, F. Hoffmann-La Roche Ltd, Basel, Switzerland

We present “domain gauges” as a set of multivariate metrics to easily quantify and characterize, with only a few numbers, drug-induced brain fMRI activation patterns. The gauges have been calibrated based on different classes (“domains”) of marketed or validated reference drugs. Capitalizing on our in-house database with an unprecedented wealth of standardized rat-brain perfusion fMRI data, we exemplarily defined and applied 3 domain gauges to demonstrate that our approach represents a valuable analytical tool to reveal drug-related neuronal circuits and concisely assess brain activation patterns, thereby facilitating compound characterization and decision making in drug discovery.

Abu Bakar Ali Asad¹, Stephanie Seah¹, Richard Baumgartner², Dai Feng², Andres Jensen¹, Brian Henry¹, Andrea Houghton³, Jeffrey Evelhoch⁴, and Chih-Liang Chin^1
1Translational Medicine Research Centre, MSD, Singapore, Singapore, ²Biometrics Research, Biostatistics and Research Decision Sciences, Merck & Co Inc, Rahway, New Jersey, United States, ³Neuroscience, Merck & Co Inc, West Point, Pennsylvania, United States, ⁴Imaging, Merck & Co Inc, West Point, Pennsylvania, United States

To bridge preclinical research and clinical investigation, we established a capsaicin pain heat fMRI model in non-human primate with heat/42 °C stimuli. We hypothesize potentiation in heat-induced cortical activation in response to capsaicin application at the forearm will highlight the ‘pain matrix’ depicted in human pain fMRI experiments. Our results indicate that group comparisons of brain activation between pre- and post-capsaicin application show significant increases in BOLD signals at the frontal, cingulate, precentral and postcentral gyrus, and cerebellum (paired t-test, p<0.002, n=8). Our data provide insights into differentiating brain regions involved with pain-like responses or thermal sensation.

Alberto Galbusera¹, Angelo Bifone¹, and Alessandro Gozzi^1
1MRI Laboratory, Istituto Italiano di Tecnologia, Centre for Neuroscience and Cognitive Sciences, Rovereto, Trento, Italy

Recent studies have highlighted the possibility of exploiting intranasal administration as non invasive method for direct delivery of large therapeutics and neuropeptides to the brain. However, the dynamics and neurofunctional substrates engaged upon intranasal administration of therapeutic agents remain unexplored. We used fMRI in the mouse to characterise the substrates activated by nasal administration of oxytocin and vasopressin, two neuroactive peptides with therapeutic potential. The two peptides elicited robust dicothomic functional responses, with significant cortico-hippocampal activation for oxytocin, and widespread cortical deactivation for vasopressin. These results support the use of fMRI to investigate the functional substrates recruited by intranasally-delivered biologicals.

Patrick P. Gao^1,2, Shu-Juan Fan^1,2, Jevin W. Zhang^1,2, Iris Y. Zhou^1,2, Joe S. Cheng^1,2, Yuqi Deng², Dan H. Sanes³, and Ed X. Wu^1,2
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, HKSAR, China, ²Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, HKSAR, China, ³Department of Biology, New York University, New York, NY, United States

Descending projections from cerebral cortex to subcortical nuclei are among the largest pathways in the brain, suggesting that they are important for subcortical processing. In the auditory system, the cortical output targets multiple subcortical nuclei. However, compared with the ascending pathways, the functional roles of these corticofugal projections are poorly understood. Non-invasive fMRI is well suited to investigating this question because it can assess activities in multiple nuclei simultaneously. In this study, bSSFP fMRI was applied to study the effect of bilateral auditory cortex ablation on the responses to sound pressure level (SPL) change in low subcortical target nuclei.

Alessio Fracasso¹, Natalia Petridou², and Serge O Dumoulin^1
1Experimental Psychology, Helmholtz Institute, Utrecht University, Utrecht, Netherlands, ²Radiology, Imaging Division, UMC Utrecht, Utrecht, Netherlands

A goal of laminar imaging is to infer the origin of signals by functional laminar profiles. We used a visual paradigm that increases responses in the contralateral hemisphere driven by retino-cortical sources and decreases responses in the ipsilateral hemisphere driven by inter-hemispheric sources. GE-BOLD data were acquired at 0.55³mm resolution. Laminar profiles were sorted according to linear trend strength towards the pial surface. Profiles showing a weak linear trend revealed a peak or dip at the middle layers when driven by retino-cortical or inter-hemispheric sources respectively. These profiles demonstrate the ability of laminar imaging to infer the origin of signals.

Halleh Ghaderi^1,2, Abbas F. Sadikot^3,4, and G. Bruce Pike^5,6
1Biomedical Engineering, McGill University, Montreal, Quebec, Canada, ²Brain Imaging Center, Montreal Neurological Institute, Montreal, Quebec, Canada, ³Neurology and Neurosurgery, McGill University, Quebec, Canada, ⁴Brain Imaging Center, Montreal Neurological Institute, Quebec, Canada,⁵Biomedical Engineering, McGill University, Quebec, Canada, ⁶Montreal Neurological Institute, Quebec, Canada

This study examines the variability in the location of task-based fMRI activation and compares these results with resting-state fMRI based target localization. The result suggests that using a predefined anatomical atlas coordinates for surgical targeting may not be accurate. But, using a combination of resting-state and task-based fMRI can reliably map somatotopy in the thalamus and somatosensory cortex and hence improve target identification for neurosurgery of deep brain structures.