BIDC Image Analysis Contributed Publications 
The BIDC routinely contributes to image analysis projects across the imaging spectrum.  

Whether you want to stop in for a brief question, schedule a time to pick our brains on a solution to your challenge, or set up a formal collaboration, our door is always open.

Below are some recent publications in which we worked as collaborators in the data analysis along with a brief desciption of the analysis difficulties and the tools and methods devised to reach the analysis goals.

Tcell:BMDC Immunological Synapse. Adapted with permission: DOI: 10.1126/science.aal3118
“Visualizing dynamic microvillar search and stabilization during ligand detection by T cells.” En Cai, Kyle Marchuk, Peter Beemiller, Casey Beppler, Matthew G. Rubashkin, Valerie M. Weaver, Audrey Gerard, Tsung-Li Liu, Bi-Chang Chen, Eric Betzi, Frederic Bartumeus and Matthew F. Krummel. 2017. Science 356(6338): 4749-4754. [10.1126/science.aal3118].
  • Finding the "search and find" dynamics of tcell microvillar protrusions in a live 4D environment is difficult for a few reasons. A major difficulty is isolating protrusion movement from the overall movement of a cell; a tcell can undergo 3D translational and rotational movement, which will convolute the motion of the protrusions. An additional difficulty lies in the movements of the protrusions themselves; protrusions will form, dissolve, merge, and also re-emerge making intensity based tracking methods hard to implement.
  • The eventual solutions were to first "stabilize" the cell within the field of view in 3D through time. This was done by first making surfaces in Imaris, which was used to create a binary mask of the isolated cell. An intensity unweighted center of mass of the cell was then found and placed in the center of the field of view for each timepoint. The protrusion dynamics could then be tackled; a local surface environment was used as a mask and the protrusions were converted to binary. The time it took for protrusions to search the local environment could then be calculated.

“Insights from imaging the implaying embryo and the uterine environment in three dimensions.” Ripla Arora, Adam Fries, Karina Oelerich, Kyle Marchuk, Khalida Sabeur, Linda C. Guidice and Diana J. Laird. 2016. Development 143(24): 4749-4754. [10.1242/dev.144386].
  • Ripla Arora (formerly from the Laird Lab, presently faculty at Michigan State University) approached us with the following problem: she wanted to characterize how the ovarian lumen folds as a function of time with respect to egg implantation. 3D surfaces of the lumen were generated by Imaris using confocal images, and a Matlab script provided by Bitplane was intended to quantify the curvature of the surface. The problem was two-fold (no pun intended): first, the script could only calculate local curvature, and since the lumen was so large, it was quantified as being flat; second, the compuatation time was extremely excessive (on the order of a day for a lumen segment). 
  • Adam Fries and Kyle Marchuk solved this problem by altering the Matlab script in two ways: first, to reduce the resolution of the surface generated by Imaris; and second, to include a much larger number of neighboring surface positions for the curvature calculation. This alteration to the code created the desired quantification; a global characterization of the surface curvature for about an hour of computation time.