Are we accepting new trainees? 

Yes! We welcome inquiries from candidates at all levels with a strong programming and/or machine learning background and prior training (or very strong interest) in developing and/or using cutting edge methods for studying cerebral cortex in humans and nonhuman primates.

Why analyze your neuroimaging data on cortical surfaces?

Because it is the best way to respect the topology of the convoluted cortical surface! Surface-based analysis enables 3-fold better spatial localization of findings in the cerebral cortex than traditional volume-based analysis approaches enabling much more reproducible neuroscience. See Van Essen et al. (1998), Glasser et al.. (Nature Neuroscience, 2016), and Coalson et al. (2018) for detailed explanations.

What is the HCP-style neuroimaging approach?

In broad strokes, ‘HCP-style’ neuroimaging data is multi-modal MRI data that has been acquired at high spatial and temporal resolution and on as many participants as is feasible, has been preprocessed to minimize spatial distortions and effects of motion along with minimizing temporal artifacts related to head motion and physiology while avoiding spatial blurring or otherwise changing the character of the signal of interest, is analyzed in a coordinate space appropriate for each brain structure and in relation to a functionally relevant parcellations, and whose data has been shared publicly and extensively analyzed data (see Glasser et al.. (Nature Neuroscience, 2016), for example on the BALSA study results repository (https://balsa.wustl.edu).

What is HCP-style neuroimaging data?

In humans, for structural MRI this means 3D T1w and T2w (or FLAIR) images of 0.8mm isotropic resolution or better to accurately map thin cortical areas (of as low as 1.6mm thickness) ideally using internal motion correction navigators with optimized tissue contrast parameters. For functional MRI this means multi-band gradient echo EPI fMRI of less than 2.6mm isotropic spatial resolution (mean cortical thickness) and a TR of less than 1 second (to best enable denoising) together with matched phase reversed spin echo EPI images for distortion and intensity bias field correction. For diffusion MRI this means using multi-band single refocusing spin echo EPI to maximize spatial resolution, diffusion weighting, and angular resolution while minimizing TE to the extent possible for a given amount of time and scanner with at least phase encoding direction reversed b0 images for distortion correction. All these sequences will benefit from vendor specific intensity bias correction (e.g. Siemens Pre-scan normalize). Non-human primates benefit from appropriately scaled down spatial resolutions (e.g. 0.5mm isotropic structural and 1.25mm isotropic functional for macaques). Detailed protocols are available here: https://www.humanconnectome.org/hcp-protocols