Smartphone-based community screening for eye disease in rural India

印度农村地区基于智能手机的眼病社区筛查

• 批准号: 10704721
• 负责人: Kunal S Parikh
• 金额: $ 18.34万
• 依托单位: JOHNS HOPKINS UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-30 至 2024-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10704721
• 关键词: Algorithms; Anterior; Anterior eyeball segment structure; Artificial Intelligence; Award; Biomedical Engineering; Biometry; Blindness; Calendar; Caring; Cataract; Cellular Phone; Clinical Data; Collaborations; Communities; Community Health Aides; Computer software; Continuity of Patient Care; Corneal Opacity; Data; Data Collection; Data Set; Dependence; Development; Devices; Diagnosis; Diagnostic; Disease; Ensure; Epidemiology; Equipment; Eye; Eye diseases; Failure; Geography; Health; Health Technology; Health system; Healthcare; Hospitals; Human Resources; Image; Improve Access; India; Information Technology; Infrastructure; Intervention; Machine Learning; Modeling; Ophthalmologist; Ophthalmology; Outcome; Patients; Pattern; Performance; Periodicals; Persons; Phase; Public Health Practice; Quality Control; Refractive Errors; Research; Research Personnel; Resource-limited setting; Rural; Specialist; Standardization; System; Time; Training; Translating; Universities; Validation; Visually Impaired Persons; Vulnerable Populations; acceptability and feasibility; algorithm training; biomedical referral center; blind; care delivery; care providers; collaborative trial; cost; cost effectiveness; deep neural network; design; effectiveness evaluation; follow-up; improved; innovative technologies; interoperability; low and middle-income countries; mHealth; machine learning algorithm; member; novel; ophthalmic examination; patient screening; prototype; remote communities; remote grading; rural patients; rural setting; screening; skills; smartphone based device; user centered design; user-friendly

PROJECT SUMMARY This is an application for an R21/R33 award titled “Smartphone-based community screening of anterior eye diseases in rural India.” The investigators from Johns Hopkins University and Aravind Eye Hospital have diverse expertise in ophthalmology, biomedical engineering, machine learning, epidemiology, biostatistics, and use of mobile health technology in low-resource settings. 90% of the world’s 275 million blind and visually impaired people live in low- and middle-income countries (LMICs). Cataract, refractive error, corneal opacities, and other anterior eye diseases account for the majority of global blindness. In rural LMIC settings, lack of access to highly trained eye care providers such as ophthalmologists is a key barrier to timely diagnosis and treatment for anterior eye diseases. Periodic “eye camp” screenings performed by highly trained ophthalmologists have been the mainstay of rural eye disease screening for many decades but have several limitations including lack of access to trained ophthalmologists, failure to reach the most remote communities, lack of consistent calendar coverage beyond eye camp dates, and high costs due to equipment, personnel, and community publicity. We propose development, validation, and implementation of a novel smartphone based device for community health worker (CHW)-led screening, diagnosis, and referral for eye diseases in rural LMIC settings. We will develop this smartphone platform using iterative prototyping and user-centered design approaches. After demonstrating feasibility, we will evaluate the diagnostic validity of CHW-led screenings using the smartphone platform compared to traditional in-person eye camp exams by an ophthalmologist. After collecting sufficient data and images using the platform, we will design, validate, and implement machine learning algorithms to permit real- time diagnosis and referral decisions by CHWs without dependence on ophthalmologists. This project will perform detailed data collection regarding referral patterns, loss to follow-up, cost-effectiveness, and differential outcomes among vulnerable groups in order to enable more targeted health interventions. This project is intended to overcome key longstanding geographic, financial, operational, and human resource constraints to eye screening while ensuring diagnostic validity, quality control, and interoperability with existing health system infrastructure. Our scalable approach has potential to transform global eye care delivery in low-resource settings. This collaboration will also result in permanent improvements in mobile health and research capacity at Aravind Eye Hospital.

项目摘要 这是R21/R33奖的申请，标题为“基于智能手机的前眼社区筛查 印度农村地区的疾病。 眼科，生物医学工程，机器学习，流行病学，生物统计学的专业知识和使用 低资源环境中的移动健康技术。全球2.75亿盲人和视力障碍的90％ 人们生活在低收入和中等收入国家（LMIC）。白内障，折射率，角膜骨质和其他 前眼疾病占全球失明的大部分。在粗糙的LMIC环境中，无法获得高度访问 训练有素的眼保健提供者（例如眼科医生）是及时诊断和治疗前部治疗的关键障碍 眼病。训练有素的眼科医生进行的定期“眼训练营”放映是 几十年来的粗糙眼病筛查中流tay柱，但有几个局限性 对于训练有素的眼科医生，未能到达最偏远的社区，缺乏一致的日历覆盖范围 超越眼营的日期，以及由于设备，人员和社区宣传而引起的高昂成本。我们建议 为社区卫生工作者开发，验证和实施基于智能手机的新型设备 （CHW）在粗糙的LMIC环境下进行眼科疾病的筛查，诊断和转诊。我们将开发这个 智能手机平台使用迭代原型制作和以用户为中心的设计方法。演示后 可行性，我们将使用智能手机平台评估CHW LED筛查的诊断有效性 与眼科医生的传统人工眼营检查相比。收集足够的数据后 使用平台的图像，我们将设计，验证和实施机器学习算法，以允许实现 CHW的时间诊断和推荐决定，而无需依赖眼科医生。这个项目将 执行有关推荐模式，损失的详细数据收集，随访，成本效益和差异 弱势群体之间的成果，以实现更多针对性的健康干预措施。这个项目是 打算克服关键的长期地理，财务，运营和人力资源限制 眼睛筛查，同时确保诊断有效性，质量控制以及与现有卫生系统的互操作性 基础设施。我们可扩展的方法有可能在低资源环境中转变全球眼保健。 这种合作还将导致Aravind的移动健康和研究能力永久改善 眼科医院。