Presenter: Vern Hart
Authors: Rylee Russell, Tyler Daynes, Jackson Wilde, Matt Phillips, Ellie Evans, Tyson Hoyt, Jeremy Tait, Jackson Aubrey, Harrison Welch, CJ Winward, Spencer Rasmussen, Brandon Jolley, Shritha Gummadi, Matt Barton, Clayton Rawson, Vern Hart
Faculty Advisor: Vern Hart
Institution: Utah Valley University
Traditional optical microscopy exhibits several limitations, most notably the classical diffraction limit that prevents the resolution of objects smaller than ~0.5 microns. However, many of the intracellular changes that occur during the earliest indications of cancer manifest on scales of less than one micron. As such, we demonstrate a low-cost on-axis digital holographic microscopy (DHM) system in which scattered light from a visible laser diode, the phase of which has been shifted by a sample, is made to interfere with an incident plane wave. This effect produces holographic patterns that can be measured using an optical beam profiler, from which amplitude and phase information can be acquired for optical path length calculations. This information, a combination of thickness and variations in the index of refraction for a sample, is simply not accessible in visible microscopy images, which only provide attenuation measurements that differ very little between cells. Experimental results will be presented for adherent mammalian cancer cells (BHK and PANC-1) serving as the imaging targets, with amplitude profiles reconstructed via the angular spectrum method. Unwrapped phase images will also be presented, in addition to early instance segmentation results involving the use of a deep neural network for identifying trace quantities of target cells in a mixed sample.