In the previous cycle of the P41 we succeeded in developing an integrated registration framework that combined volumetric information with surface-based features to yield a single coordinate system that provides unsurpassed accuracy for the alignment of structural features of the human brain^'^. In addition to developing this Combined Volume and Surface (CVS) registration tool, we have distributed it to our service users and collaborators, and integrated it into FreeSurfer (FS) so that it is now available to our wider user community of more than 13,000 scientists and clinicians. In response to new and existing Collaborations we have established a set of new aims that will greatly facilitate the research of our collaborative and service projects (CPs and SPs). One group of aims involves developing tools to replace the time-intensive (and hence motion-sensitive) acquisition of functional localizers to define multiple cortical regions (e.g. the fusiform face area FFA, rTPJ, etc.). For these CPs we propose to develop tools for predicting the location of functionally defined regions-of-interest (ROIs) in the cortex from a single resting state fMRI and/or diffusion-weighted MRI in conjunction with cortical folding patterns (CPs Buckner, Gabrieli, Saxe). In this cycle we have also established new collaborations that motivate us to extend our in vivo probabilistic architectonic labeling to frontal regions important in depression and anxiety disorders (CPs Haber, Mayberg and Milad). The architectonic features that define these borders are too subtle to detect with even lOO?m ex vivo MRI, and we have therefore begun a synergistic aim with Project 4 to develop Optical Coherence Tomography (OCT) in conjunction with ex vivo MRI to define architectonic boundaries in the human cortex, focusing on cortical areas 25 and 32. Other new CPs (Gollub, Brown) have encouraged us to extend our anatomical labeling to brainstem and deep brain nuclei that are important in pain processing and arousal, as well as neurodegenerative diseases such as Parkinson's, and for the successful implantation of deep brain stimulators (DBS, CPs Haber and Mayberg). Here we will work with Project 2 to acquire high-resolution high contrast-to-noise ratio (CNR) images and use optimal image restoration techniques to reduce the acquisition time to a feasible protocol. These restoration techniques can also be used to synthesize images across different contrast, with important clinical applications (CP Murphy). The end result of these developments will be a more accurate cross-subject coordinate system and robust, automated labeling of a collection of brain regions that are critical for making progress in an array of clinically and scientifically important problems.
Showing the most recent 10 out of 300 publications
