Abstract
The recent increase in the use of high field MR systems is accompanied by a demand for acquisition techniques and coil systems that can take advantage of increased power and accuracy without being susceptible to increased noise. Physical location and anatomical complexity of targeted regions must be considered when attempting to image deeper structures with small nuclei and/or complex cytoarchitechtonics (i.e. small microvasculature and deep nuclei), such as the brainstem and the cerebellum (Cb). Once these obstacles are overcome, the concomitant increase in signal strength at higher field strength should allow for faster acquisition of MR images. Here we show that it is technically feasible to quickly and accurately detect blood oxygen level dependent (BOLD) signal changes and obtain anatomical images of Cb at high spatial resolutions in individual subjects at 7 Tesla in a single one-hour session. Images were obtained using two high-density multi-element surface coils (32 channels in total) placed beneath the head at the level of Cb, two channel transmission, and three-dimensional sensitivity encoded (3D, SENSE) acquisitions to investigate sensorimotor activations in Cb. Two classic sensorimotor tasks were used to detect Cb activations. BOLD signal changes during motor activity resulted in concentrated clusters of activity within the Cb lobules associated with each task, observed consistently and independently in each subject: Oculomotor vermis (VI/VII) and CrusI/II for pro- and anti-saccades; ipsilateral hemispheres IV-VI for finger tapping; and topographical separation of eye- and hand- activations in hemispheres VI and VIIb/VIII. Though fast temporal resolution was not attempted here, these functional patches of highly specific BOLD signal changes may reflect small-scale shunting of blood in the microvasculature of Cb. The observed improvements in acquisition time and signal detection are ideal for individualized investigations such as differentiation of functional zones prior to surgery.
Highlights
Cerebellar FunctionThe cerebellum (Cb) has a uniform architecture throughout and is divided into the cerebellar vermis (v) along the medial portion and cerebellar hemispheres (h) laterally with the paravermis located between the two; these regions are divided into ten lobules, arranged dorsoventrally (Fig 1a)
Accuracy of placement was confirmed with a scout scan and subjects were repositioned if Cb was not within the field of view (FOV) of the surface coils. (For an example of data of one subject acquired when the Cb was below the FOV of the coils, see S1 Fig) RF transmit phases were adjusted separately to homogenize the B1 field around Cb and the B0 field was shimmed separately on the FOV using pre-defined shim tools built in house for both procedures [61]; these shimming parameters were applied to all subsequent acquisitions, including the coil sensitivity profile acquisition
The data of one subject were acquired when the Cb was below the FOV of the coils and these data are not displayed in the figures of the main text; see S2 Fig for complementary panels of structural and functional images from this subject
Summary
Cerebellar FunctionThe cerebellum (Cb) has a uniform architecture throughout and is divided into the cerebellar vermis (v) along the medial portion and cerebellar hemispheres (h) laterally with the paravermis (or intermediate zone) located between the two; these regions are divided into ten lobules, arranged dorsoventrally (Fig 1a). Many studies have implied that (lateral) Cb cortex is critical to cognitive or goal-directed neocortical processes involved in controlling volitional eye movements and/or appropriate suppression of reflexive eye movements, such as memory-guided saccades, anti-saccades, and saccade adaptation [24,25,26,27,28]. This viewpoint can be challenged, as attention and eye movements are an integral part of most studies on cognition [29,30,31,32]. Though it has been difficult to pinpoint the location(s) of different aspects of cognitive processes within Cb, relating the cognitive planning components of volitional movements (versus reflexive movements [33]) and/or motor learning are a good starting point
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
