Point of care diagnostic for sickle cell disease

项目摘要 镰状细胞病（SCD）是一种严重的血液病，见于撒哈拉以南非洲和美国，影响 占非洲新生儿总数的3%。在美国，这种疾病影响了10万美国人，但 非洲裔美国人不成比例地承担着每365名非洲裔美国人中就有一人受到影响。SCD 患者经历血管闭塞危象（VOC）作为急性发作的疼痛。目前，治疗SCD的唯一方法是 骨髓移植，但这一过程的成本和困难导致SCD患者往往选择 而是通过治疗和常规监测来控制疾病。然而，目前没有客观的 测量或实验室测试，以确定VOC和相关的疼痛严重程度。常规监测红细胞（RBC） 然而，健康可以为避免VOC提供一个窗口，并能够更彻底地定量分析 治疗干预。 本项目的目标是创建一个原型设备，适用于监测SCD的护理点 通过推进快速、自动化、高通量红细胞（RBC）成像的新技术来帮助患者。 该方法是基于定量相位成像（QPI），以创建一个全息图的每一个单独的红细胞在一个 sample.然后可以使用机器学习来评估这些成像数据中镰状细胞的比例。 此外，由于每个样本中多达10^6个RBC被成像，因此将获得大量数据，适合于 开发可以揭示更多信息的人工智能算法。值得注意的是，该装置结构紧凑， 成本，这表明有很大的潜力转化为现场护理设备。在这个项目中，我们将 我们的技术已被证明具有高通量成像能力，我们希望将其转化为原型 适用于临床试验的器械。关键的开发步骤包括成像设备的开发 以实现坚固、紧凑的外形，推进携带RBC的微流体装置的设计， 并创建新的和改进的算法，以实现高通量分割和分析所获得的 成像数据。输出将是一个原型设备，用于评估个人的红细胞健康的基础上使用 可以在本地进行分析，以便更快地做出治疗决定。

Project Summary Sickle cell disease (SCD) is severe hematological disease seen in Sub-Saharan Africa and the U.S., affecting up to 3% of the newborn population Africa. In the US, the disease affects 100,000 Americans, but is disproportionately borne by African Americans with one out of every 365 African-Americans affected. SCD patients experience vaso-occlusive crises (VOCs) as acute bouts of pain. Currently, the only cure for SCD is a bone marrow transplant but the cost and difficulty of this procedure results in SCD patients often choosing to instead manage their disease with treatments and routine monitoring. However, there is currently no objective measure or lab test to determine VOCs and associated pain severity. Routine monitoring of red blood cell (RBC) health, however, could provide a window for avoiding VOCs and enable more thorough, quantitative analysis of therapeutic interventions. The goal of this project is to create a prototype device suitable for use at the point of care for monitoring SCD patients by advancing a novel technology for rapid, automated, high throughput red blood cell (RBC) imaging. The approach is based on quantitative phase imaging (QPI) to create a hologram of every individual RBC in a sample. These imaging data can then be evaluated for the proportion of sickled cells using machine learning. Further, since up to 10^6 RBCs are imaged in each sample, a large volume of data will be obtained, suitable for developing AI algorithms that can reveal additional information. Significantly, the device is compact and low cost, indicating that there is significant potential for translation to a point of care device. In this project, we will seek to translate our technology, with demonstrated high throughput imaging capabilities, to realize a prototype device suitable for use in clinical trials. They key development steps include development of the imaging device to enable a robust, compact form factor, advancing the design of the microfluidic device that carries the RBC's and creating new and improved algorithms to enable high throughput segmentation and analysis of the obtained imaging data. The output will be a prototype device for assessing the RBC health of individuals based using a small blood draw which can be analyzed locally to enable treatment decisions more quickly.