Researchers at Georgia Tech have developed a new algorithm for identifying and tracking cell membrane locations. This has the potential to automate electrophysiological measurements such as patch clamping, the “gold standard” of single-cell electrophysiology. The procedure leverages DIC microscopy to examine in vitro brain slices. Data is processed through a novel deconvolution algorithm formulated as a regularized least-squares optimization with the goal of providing automated visual tracking of a target cell’s membrane to guide automated electrophysiology systems (e.g., robotic patch clamping systems).

Importantly, this unique process can sift through high levels of organic tissue interference and can operate in environments with moving tissue. The algorithm filters out noise in the current image and then deconvolves and segments that image to determine cell edges based on the data from images earlier in the sequence. The final stage is a segmentation based on that deconvolution, which is performed with simple thresholding.