This is the shared lab wiki for the VisCog group at the University of Washington. We study the cortical basis of sensory function in both normal and atypical individuals. Our work relies on multiple methodologies, including behavioral measurements, functional magnetic resonance imaging, magnetic resonance spectroscopy (measuring neurochemicals in the human brain), and we generally try to use computational methods to tie it all together with a nice little equation that's not too complicated. For more about who we are and what we do, please visit our lab webpage.
Website that allows you to import data from graphs. Geoff also has a grabdata.m function in the UW_toolbox.
Basically there are two ways to run big processes - the viscog server and local multicore machines. Recent bunch results:
LU FFT ODE Sparse 2-D 3D
Dell XPS (laptop 4 cores) 0.1176 0.0875 0.0134 0.1096 0.5427 0.5562 VISCOG .1135 .1608 .0970 .2795 3.2305 3.7418 Mark (Ione’s Office) 12-cores 0.0881 0.0944 0.0150 0.0754 0.4483 0.4302
More than 20 million Americans aged 18 and older have experienced vision loss, and rates of vision loss are expected to double by 2030 as the nation’s population ages. A variety of sight recovery therapies are being developed by companies around the world, offering new hope for people who are blind. But little is known about what the world will look like to patients who undergo such procedures. We have developed a computer model of what someone with restored vision might see. This computer model creates a ‘virtual patient’ who can give scientists and patients into the vision that these prostheses might provide. Python. http://uwescience.github.io/pulse2percept/
Code for fitting population receptive fields (Dumoulin & Wandell, 2008). Includes most of the stuff you might want to do: estimating the HDR, compressive spatial summation. Unlike some other kinds of pRF code it's designed to be pretty independent of your analysis platform. Matlab. https://github.com/kellychang4/pRF.
This retinotopic code was written to be extremely flexible and serve a number of purposes. These include 1. Including code for multifocal (Vanni Neuroimage 2005) as well as drifting bars 2. Code is written has having a background image windowed by an aperture. Allows you to have anything from a flickering checkerboard to a pixar movie in the background 3. Generates a binary mask, useful as input into prf mapping 4. Includes a very attentionally demanding Simon task 5. Easy to change the size/position of fixation spot 6. Can also create a 'fake patient' by a) jittering the fixation to match fixation instability (you can even read in a real patients eye-trace) and adds a scotoma. Matlab & Psychtoolbox. https://github.com/ionefine/retinotopic-mapping
This is a handy set of Matlab functions that do all sorts of useful things for people in vision science and neuroimaging, including functions for psychometric curve estimation and plotting. A list of functions is here. The depository is not yet on github.