Studying pseudogout using natural language processing and novel imaging approaches

项目概要/摘要（来自受资助的PANK 23申请） 候选人：Tedeschi博士是哈佛医学院（HMS）的医学讲师， 临床科学部（SCS）的流变学、免疫学和过敏科医生， 布里格姆妇女医院（BWH）她获得了哈佛T.H.陈公学校 健康（HSPH）。她的10个第一作者手稿，两个BWH赠款，和基金会奖 生产力和对研究的承诺。她组建了一支经验丰富的导师团队， 合作者，由丹尼尔所罗门博士（主要导师）和凯瑟琳廖博士和卡伦Costenbader领导 （共同导师），指导她在自然语言处理和机器学习方法方面的培训， 研究、分析关联的电子病历（EMR）和医疗保险索赔数据， 先进的成像模式。在HSPH和HMS的重点课程将补充她的经验， 通过她提出的假痛风危险因素和长期结果的研究获得的收益。培训在双- 水晶关节炎的能量CT和超声解释将通过一对一的会议获得。她 长期的职业目标是成为一个独立的以病人为导向的研究者，专注于假性痛风。 环境：Tedeschi博士从她的部门承诺>75%的保护时间用于研究， 在K23奖励期间的职业发展活动。该司及其主要导师的支持 研究基金将补充她的工资和项目相关费用。SCS是一个合作的临床 流变学部的一个研究小组，拥有广泛的基础设施，包括VERITY 生物信息学核心（NIH-P30-AR 072577，PI：所罗门），将为 建议的研究。此外，BWH关节炎中心是全国最大的促进学科之一， BWH肌肉骨骼成像部门拥有最先进的设备和专业知识 双能量CT在结晶性关节炎中的应用课程在HSPH和HMS，毗邻BWH，将提供 Tedeschi博士成为独立调查员所需的培训。研究：Tedeschi博士 长期目标是预防和减少假性痛风的发病率，假性痛风是一种未充分研究的疼痛性晶体 关节炎，影响800 - 1000万美国人。她将使用自然语言处理和机器学习 方法，以增强识别EMR数据中的假痛风的算法。她将研究风险因素， 假痛风的长期结果，利用Partners HealthCare中包含的大量信息， EMR数据和Medicare索赔数据，并将获得使用链接数据集的经验。特德斯基医生会 招募患有假性痛风和其他类型的单关节炎和少关节炎的受试者， 双能CT扫描、肌肉骨骼超声和X线检查用于识别假性痛风。 她提出的K23项目将导致在R 01应用程序中利用手稿和数据，重点是 假痛风在授予期间，导致独立作为一个以病人为导向的调查。

PROJECT SUMMARY/ABSTRACT (FROM FUNDED PARENT K23 APPLICATION) Candidate: Dr. Tedeschi is an Instructor in Medicine at Harvard Medical School (HMS) and Associate Physician in the Division of Rheumatology, Immunology and Allergy’s Section of Clinical Sciences (SCS) at Brigham and Women’s Hospital (BWH). She received an MPH from the Harvard T.H. Chan School of Public Health (HSPH). Her 10 first-author manuscripts, two BWH grants, and foundation award exemplify her productivity and commitment to research. She has assembled an experienced team of mentors and collaborators, led by Dr. Daniel Solomon (primary mentor) and Drs. Katherine Liao and Karen Costenbader (co-mentors), to guide her training in natural language processing and machine learning approaches for clinical research, analysis of linked electronic medical record (EMR) and Medicare claims data, and interpretation of advanced imaging modalities. Focused coursework at HSPH and HMS will complement the experience she gains through her proposed studies of pseudogout risk factors and long-term outcomes. Training in dual- energy CT and ultrasound interpretation for crystalline arthritis will be obtained via one-on-one sessions. Her long-term career goal is to become an independent patient-oriented investigator focused on pseudogout. Environment: Dr. Tedeschi has a commitment from her Division for >75% protected time for research and career development activities during the K23 award period. Support from the Division and her primary mentor’s research funds will supplement her salary and project-related expenses. The SCS, a collaborative clinical research group in the Division of Rheumatology, has extensive infrastructure including the VERITY Bioinformatics Core (NIH-P30-AR072577, PI: Solomon) that will provide resources and expertise for the proposed studies. In addition, the BWH Arthritis Center is one of the largest nationally, facilitating subject recruitment, and the BWH Division of Musculoskeletal Imaging has state-of-the-art equipment and expertise applying dual-energy CT in crystalline arthritis. Coursework at HSPH and HMS, adjacent to BWH, will provide training necessary for Dr. Tedeschi’s development into an independent investigator. Research: Dr. Tedeschi’s long-term objective is to prevent and reduce morbidity from pseudogout, an understudied, painful crystalline arthritis that affects 8-10 million Americans. She will use natural language processing and machine learning approaches to enhance an algorithm for identifying pseudogout in EMR data. She will study risk factors for and long-term outcomes in pseudogout, harnessing vast amounts of information contained in Partners HealthCare EMR data and Medicare claims data, and will gain experience working with linked datasets. Dr. Tedeschi will recruit subjects with pseudogout and other types of mono- and oligoarthritis to test and compare the performance of dual-energy CT scanning, musculoskeletal ultrasound, and x-ray for identifying pseudogout. Her proposed K23 projects will lead to manuscripts and data to be leveraged in an R01 application focused on pseudogout during the award period, leading to independence as a patient-oriented investigator.