Automated diabetic retinopathy screening system

描述（由申请人提供）：该拟议项目的动机是，如果不引入视网膜图像的计算机辅助诊断，对糖尿病患者进行全面、广泛的视网膜病变筛查在经济上是不可行的。 美国有超过2000万人患有糖尿病，据估计，这些人中只有不到一半的人定期接受糖尿病视网膜病变筛查。 缺乏医疗保险和获得医疗保健提供者的机会对近1000万糖尿病患者构成了主要障碍。 创建一个负担得起的和可访问的解决方案，为这些糖尿病患者提供筛查服务，这对医疗保健界提出了一个重大挑战。 美国公民利用人类“读者”对每个病例进行评分的全面筛查计划将是昂贵的。 解决办法是采用类似于乳房X光检查和子宫颈抹片检查等其他医疗应用的计算机辅助技术;这将为我们的高危人群提供全面、定期的筛查。 许多研究者已经开发了特定的算法，每种算法仅检测一种特定类型的病变，例如暗病变、白色病变或血管特征。这些算法已使用单一模态（像素格式、SLO、标准眼底镜、彩色、无红色等）进行了测试。 并且每个新的摄像机可能需要对算法进行显著的重新调整。这个项目的目标是展示并验证一种全新的计算机辅助视网膜图像分级方法。 这项拟议研究的意义是双重的。首先，通过提高阅读中心的生产力，更多的高危糖尿病患者将获得这项服务，从而通过早期发现和治疗提高生活质量。第二，通过提供一个有效的，强大的计算机辅助评分系统，所有现有的阅读中心将受益于经济上通过我们的系统的效率增加。我们的系统不会取代目前的人类读者;它只是允许增加五个或更多的因素的情况下，不牺牲灵敏度和特异性的吞吐量。 公共卫生相关性：今天，约有1000万糖尿病患者没有接受年度眼科检查。如果没有这些检查，早期发现威胁视力的视网膜病变是不可能的。在西方世界，许多糖尿病患者由此导致的早期视力丧失是失明的主要原因。 没有足够数量的医疗保健专家为这一人群进行眼科检查。 没有计算机筛查，就不可能对人口进行大规模筛查。

DESCRIPTION (provided by applicant): This proposed project is motivated by the fact that comprehensive, broad-scale screening of diabetics for retinopathy is economically prohibitive without the introduction of computer-assisted diagnosis of retinal images. Over 20 million people in the U.S. have diabetes and it is estimated that less than half of those individuals are screened periodically for diabetic retinopathy. Lack of medical coverage and access to healthcare providers imposes major obstacles for nearly 10 million diabetics. Creating an affordable and accessible solution to providing screening services to these diabetics presents a significant challenge to the healthcare community. A comprehensive screening program for U.S. citizens utilizing human "readers" to grade each case would be prohibitively expensive. The solution is to implement a computer-assisted technology similar to other medical applications, such as mammograms and Pap smears; that would provide comprehensive, periodic screening of our at-risk population. Numerous investigators have developed specific algorithms, each detecting only one specific type of lesion, such as dark lesions, white lesions, or vessel characteristics. These algorithms have been tested using a single modality (pixel format, SLO, standard funduscope, color, red free, etc.) and each new camera can require significant re-tuning of the algorithms. The goal of this project is to demonstrate and then validate an entirely new approach for computer-assisted grading of retinal images. The significance of this proposed research is two-fold. First, by increasing the productivity of reading centers, a much larger population of at-risk diabetics will have access to this service, leading to improved quality of life through early detection and treatment. Second, by providing a validated, robust computer-assisted grading system, all existing reading centers would benefit economically through the added efficiency of our system. Our system does not replace current human readers; it simply allows an increase of throughput of cases by factors of five or more without sacrificing sensitivity and specificity. PUBLIC HEALTH RELEVANCE: Today, about 10 million diabetics do not receive annual eye examinations. Without these examinations early detection of vision threatening retinopathy is not possible. The resulting early loss of vision for many of these diabetics is the leading cause of blindness in the western world. There is an insufficient number of healthcare specialists to perform eye examinations for this population. Without computer-based screening, a broad-scale screening of the population will not be possible.