Hizlari gonbidatua: Luis Hernández-García. Development of Layer Specific FMRI at 3 Tesla (a work in progress)

What: Development of Layer Specific FMRI at 3 Tesla (a work in progress)

Where: Auditorium and Auditorium zoom room  (If you would like to attend to this meeting reserve at info@bcbl.eu)

Who: Luis Hernández-García. PhD, Departments of Radiology and Biomedical Engineering, University of Michigan, Ann Arbor, MI, USA

When:  Tuesday, Mar 18th at 3:00 PM

Functional MRI has proven to be an excellent tool to probe brain circuits at a macro-scale over the last three decades.  Despite the high resolution achievable by modern MRI scanners, the specificity of the detected brain activity is limited in part by the spatial resolution of the scans but, ultimately, by the nature of the BOLD response which is heavily weighted toward the veins and pial vessels.  While T2* weighted images are very sensitive to the BOLD effect, they are clearly biased toward larger draining veins and the pial surface.  On the other hand, Perfusion and Vascular Space Occupancy methods can be used instead to measure changes in blood volume at the capillary level.  These approaches have been shown to be more specific to the brain tissue being activated. However, VASO imaging and its variants are not without challenges: they are slow, SNR-inefficient and often limited to a small number of slices at a time.  

Laminar specific FMRI refers to FMRI studies that can identify the cortical layers that originate the observed activity.  Laminar FMRI is typically based on arterial blood volume imaging and requires the superior SNR achievable by ultra-high field scanners (7T and above).  Identifying which cortical layer originates activity is desirable because can differentiate between bottom-up or top-down connections, adding an important dimension to the characterization of brain connectivity.  This new level of nuance is enabling a new generation of neuroscience studies in humans at the mesoscopic scale.  Importantly, this new level of nuance could potentially transform FMRI into a clinically useful tool at the individual patient level.  Unfortunately, achieving those results on clinical scanners (1.5 and 3T) is a major challenge posed by the inherently lower SNR available at lower field.

While smaller 7T scanners for rodents are not uncommon, human 7T scanners are limited to a reduced number of institutions.   The scarcity and significantly higher cost of 7T human scanners poses a significant challenge to the field. Clearly, low field layer specific FMRI is very desirable because of the cost and availability of clinical scanners, but also because of the reduced BOLD contamination of the arterial signals at lower field strength.  In this talk, I will present progress on our project to make layer specific FMRI available and practical in clinical scanners (i.e., 3 Tesla). Our strategy consists of leveraging several advanced MRI techniques: (1) 3D volume tailored RF pulses to suppress aliasing  (2) rapid sampling of 3D k-space data using a spherical trajectory to optimize the trade-offs between SNR, speed and spatial resolution and (3) developing new alternatives to the VASO technique through the use of velocity selective pulses which will enable significantly more rapid sampling of the activation time course, than the standard IR-VASO.