SBIR Phase I: Hardware and Software Systems for High Throughput, High Cell Density Fermentation

点击翻译按钮获取中文摘要

This SBIR Phase I project proposes to develop a high throughput, imagery-based single cell analysis system. The system will be used to characterize high cell density polymicrobial samples captured during biological research. Initially, the technology will be applied to the problem of quantifying extremely low levels of microbial contamination in biological samples. The quantification and control of microbial contamination is an important challenge in industrial fermentation and in the manufacturing of biologics. Methods for rapid and sensitive detection of microbial contaminants do not exist. Traditional contamination detection methods involve time-consuming culturing of cells or manual and imprecise checks with a microscope. The proposed technology will provide automated and rapid sample characterization, allowing a speed of analysis that is 1000 times faster than current systems. This increased throughput will additionally allow lower levels of contamination to be quantified. While initially applied to research, the technology can also be used as an analytical and diagnostic tool in the medical field. For example, this analysis system could be used to search for malformed red blood cells in a human sample that might arise due to a disease like sickle cell anemia, or to determine the relative abundance and morphology of immune cells, an analysis that would help identify certain leukemias.This project will create a hardware platform and a parallelized computer vision system that is capable of fully characterizing each cell in a 1mL biological sample in under five minutes. To analyze samples taken from a fermentation process with high cell densities requires the micrography system to be capable of processing over one billion cells in the analysis window. The project aims to develop four subsystems: a high pressure microfluidic system for separating and isolating cells, an imaging system, a set of high throughput algorithms to classify the cellular imagery data and a horizontally-scalable computing platform on which to run the classifying code. The project will begin by developing a low throughput, fully functioning prototype. The efforts will then focus on developing the four primary subsystems in parallel, using knowledge gained in the prototyping efforts to guide the later work. Additional phases will focus on the integration of high throughput versions of the subsystems. The fully realized single cell micrography system will represent a leap forward in speed and specificity compared to traditional cytometry systems. With no manual intervention and in an open-ended, label-free manner, researchers will be able to conduct parts per billion-level inspection of a biological sample.