Low-cost Fourier ptychographic microscopy for neuroimaging applications

For a fluorescence microscopy system, a set-up often costs upwards of one hundred thousand dollars, if not multiple hundreds. This makes research inaccessible to any institutions except those with the highest level of funding. Our research addresses the challenges of high-magnification fluorescence microscopy by using low-cost off-the-shelf components to create a fluorescence imaging system. We use a cost-effective widefield system for neuroimaging using an OLED array via an Arduino microcontroller and a custom-made image processing algorithm. The OLEDs enable fluorescence microscopy with their green and blue emissions having significant overlap with the absorption spectrums for the Alexa Fluor 532 and FITC fluorescent probes. We use this system for Fourier ptychography and leverage image processing algorithms to achieve high-resolution images by combining low-resolution images within the Fourier plane. A phase retrieval algorithm and a regression model are utilized to find optimal coefficients for our transform matrices. This project is promising to democratize neuroimaging, especially in resource-constrained settings with affordable costs and portability. It is small-scale, 6'x8'x12', and compatible with any MATLAB system, thereby allowing even further reach into under-served areas.