High-throughput quantitative phase imaging for early-stage cancer cell screening with optical time-stretch technique

This project aims at applying a recently developed ultrafast imaging technique, namely quantitative phase time-stretch imaging, to early-stage cancer cell screening. At the beginning of cancer development, the occurrence of circulating tumor cells (Cancer cells that are circulating in blood stream) is extremely rare--- less than 1 in 1,000,000 cells. To identify abnormal cells, it requires fast imaging system at >10,000cells/second throughput, which is not supported by current microscope techniques. Our technique, called optical time-stretch, could fulfill the above requirement which has the potential to assist early-stage cancer diagnosis.
In our biweekly experiment, billions of image pixels (i.e. ~1–10 Gigabytes of raw data) would be collected in one batch. In order to analyze these big data in a timely manner, high-throughput computing infrastructure with highly-parallel image processing capability is required.