Video Assisted Islet Cell Isolation System

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Islet cell transplantation offers the hope of a cure for the more than 1 million Americans who are type I (insulin dependant) diabetics. However, current transplantation outcomes are highly dependent on the skill of the physicians, scientists, and technicians involved in the process of isolating and purifying islets from donor organs. To help reduce the large variabilities and manpower inherent in the current methods, Emerging Concepts, Inc., in collaboration with The University of Cincinnati and Sheet Dynamics, Ltd., is developing an islet cell isolation system to automate the major steps in the isolation process. The system holds promise for standardizing and reducing process variability, and will allow for the optimization of yield, viability, and purity of the islet cells. Currently, the system automates multiple key steps in the process, but still requires a trained technician to determine one of the critical process variables, namely, the end point of the cell dissociation stage. A high quality, automated video analysis system could significantly improve the inherent process variability by improving this key decision point. Therefore, the focus of this Phase I feasibility study is to develop the necessary digital image processing and analysis software needed for such a video system. This proposal has three specific aims: (1) develop a database of cell images and surrogate outcomes during the dissociation phase of islet cell isolation; (2) develop image analysis algorithms to distinguish between islets and exocrine tissue, extract key features and metrics from the segmented images, and formulate a predictive model that relates these metrics to optimal dissociation halting times derived from the surrogate outcomes; and (3) build a virtual microscope (a web portal) and collect responses from islet cell transplantation experts operating on the database of image sequences generated in the first aim. This Phase study will be considered a success if the performance of the predictive model can be shown to be equal to or better than (a=.05) the panel of experts operating on the same data.