Developed algorithms for multi-view reconstruction of small bugs with labels using lytro cameras. Traditional MVE problems require hundred’s of images with small baselines. We digitized small bugs (from museums) and their multiple labels, using few images with wide baselines.

Developed new methods of non-parametric feature selection in lung cancer cells to better target features which could be used as classifiers. Improved the current 3D segmentation for better feature assessment. Developed tools to increase the throughput of the entire imaging & classification pipeline.

I developed a combined procedure of sample preparation, tomographic imaging and computational processing for quantification of DNA content in single cells using 3D-OPTM for early cancer diagnosis. OPTM (Optical Projection Tomographic Microscope) is a technology that can image a cell in multiple modes in 3D, providing accurate morphologic and molecular information.