SBIR Phase I: Seamlessly Integrated Methodology for Quick and Accurate Diagnosis of Digestive Diseases from Endoscopic Video

这个小企业创新研究一期项目是开发一种名为“GISentinel”的新型多维集成软件系统和技术，用于快速检测和可靠诊断胶囊内窥镜视频图像中的异常。利用所提出的算法，GISentinel能够交互式、自动地从视频胶囊图像中检测出溃疡、出血等异常情况，并使胃肠道医生能够更好地诊断异常情况。提出的混合方法将为GT医生提供更好的工具来感知解剖标志以及病变，息肉和其他异常的颜色，形状和大小，并更准确地诊断疾病的严重程度。如果成功，自动检测算法将显著减少胶囊内镜（CE）的评估时间。美国国家消化系统疾病信息中心（NDDIC）估计，美国每年有6000万到7000万人受到消化系统疾病的影响。胶囊内窥镜的市场正在快速增长。如果成功，在SBIR项目下开发的GISentinel软件系统将是一个强大的软件工具，它利用多维信息快速准确地诊断胶囊内窥镜图像中的异常。拟议的系统和技术的商业化潜力是巨大的。

This Small Business Innovative Research Phase I project is to develop a novel multidimensional integrated software system and technology called "GISentinel" for quick detection and reliable diagnosis of abnormalities from capsule endoscopic video images. Using the proposed algorithm, GISentinel is able to interactively and automatically detecting ulcers, bleeding and other abnormalities from the video capsules images and allowing GI doctors to better diagnose the abnormality. The proposed hybrid approach will provide GT doctors a better tool to perceive anatomical landmarks and the color, shape and size of lesions, polyps and other abnormalities, and to more accurately diagnose the severity of the disease. If successful, the automatic detection algorithm will significantly reduce assessment time in capsule endoscopy (CE). The National Digestive Diseases Information Clearinghouse (NDDIC) estimates that between 60 and 70 million people in the US are affected by digestive diseases each year. The market for Capsule endoscopy is a rapidly-growing. If successful, the GISentinel software system developed under this SBIR project will be a powerful software tool that utilizes multi-dimensional information to rapidly and accurately diagnose abnormalities from capsule endoscopic images. The commercialization potential of the proposed system and technology is significant.