Mentoring in Patient Oriented Research in Osteoarthritis

项目摘要 这个K24职业中期研究者奖在以患者为导向的研究（POR）将利用我现有的POR 计划，提高我和我的学员的专业发展，并支持我不断努力， 为风湿性和肌肉骨骼POR的早期临床研究者提供必要的指导 疾病（RMD），重点是骨关节炎（OA）和健康公平。尽管这是一个非常普遍的， OA是一种严重的疾病，缺乏有效的治疗方法，仍然是个人和医疗保健的巨大负担 系统我们发表了大量关于OA各方面的性别、种族和社会经济差异的文章 根据我们在两个不同人群中的经验，约翰斯顿县OA项目 （JoCoOA）和约翰斯顿县健康研究（JoCoHS）。我的POR职业生涯一直专注于OA的各个方面 流行病学，成像和生物标志物，最近强调使用机器识别表型 学习和精准医学，以改善对这种慢性疾病的研究和管理。有必要 了解人工智能的作用，包括机器学习，无论是改善或延续 健康不公平，如已知存在于OA中的不公平。这种先进的方法，如果应用得当，将 可能为OA的表型提供新的见解，并为不同人群的治疗提供信息。 目标是提高卫生公平性。此外，POR，健康公平和机器学习的交叉点是 学员的极大兴趣和RMD培训未来临床科学家的关键需求领域，这是 K24机制的目标。我们将利用JoCoOA的现有数据，JoCoOA是一项为期30年的纵向研究， 超过4000名45岁及以上的黑人和白色男性和女性，以及新的JoCoHS，一个积极注册的 队列（2019-，n~1500），包括35-70岁的西班牙裔个体， 黑色或白色。我们将利用这些丰富的数据来源和我们广泛的机构资源来解决两个问题， 围绕疼痛和症状的亚组及其与结构和功能的关联的具体目标， 注重卫生公平。在目标1中，我们将努力验证和优化深度学习算法， 不同的队列，以更好地定义影像学特征与膝关节疼痛体验之间的关联 OA。在目标2中，我们将描述JoCoHS队列中的症状表型（其具有更广泛的 可用的症状数据），并将这些与Aim 1中收集的影像学信息以及 功能和体力活动的有效评估，以确定最可能受益于 有针对性的干预措施。这个K24项目将提供关键的保护时间和资源，以促进我的 专业发展和提高我的指导能力，同时提供一个 学员的跳板，1）在个性化项目的背景下发展他们的研究技能，2）发展 大数据分析和基于机器学习的方法的关键技能，以及3）回答关键POR问题 通过多样性和健康公平透镜，在RMD和OA中开展工作。

Project summary This K24 Midcareer Investigator Award in Patient Oriented Research (POR) will leverage my existing POR program, enhance my professional development and that of my trainees, and support my ongoing efforts to provide essential mentoring to early stage clinical investigators in POR in the rheumatic and musculoskeletal diseases (RMDs), with a focus on osteoarthritis (OA) and health equity. Despite being a highly prevalent and serious disease, OA lacks effective therapies and remains a large burden on individuals and the health care system. We have published extensively on sex, race, and socioeconomic differences in various aspects of OA based on our experience with two diverse population-based cohorts, the Johnston County OA Project (JoCoOA) and the Johnston County Health Study (JoCoHS). My POR career has focused on aspects of OA epidemiology, imaging, and biomarkers, with a more recent emphasis on identifying phenotypes using machine learning and precision medicine to improve studies and management of this chronic disease. It is essential to understand the role of artificial intelligence, including machine learning, in either improving or perpetuating health inequities such as those known to exist in OA. Such advanced approaches, if applied appropriately, will likely provide new insights on phenotypes of OA, and inform therapeutics in diverse populations with the objective of improving health equity. Further, the intersection of POR, health equity, and machine learning is of great interest to trainees and an area of critical need for training future clinician scientists in RMDs, which is the goal of the K24 mechanism. We will leverage existing data from the JoCoOA, a 30-year longitudinal study of over 4000 Black and White men and women aged 45 and older, and the new JoCoHS, an actively enrolling cohort (2019-, n~1500) that includes individuals who are 35-70 years of age and who identify as Hispanic, Black, or White. We will utilize these rich data sources and our extensive institutional resources to address two specific aims around subgroups of pain and symptoms and their associations with structure and function, with a focus on health equity. In Aim 1, we will work to validate and optimize a deep learning algorithm in these diverse cohorts to better define the association between radiographic features and the pain experience in knee OA. In Aim 2, we will characterize symptomatic phenotypes in the JoCoHS cohort (which has more extensive available symptomatic data), and associate these with the radiographic information gleaned in Aim 1 as well as validated assessments of function and physical activity, to identify subgroups that could most benefit from targeted interventions. This K24 project will provide crucial protected time and resources to promote my professional development and enhance my mentoring capabilities while simultaneously providing a springboard for trainees to 1) develop their research skills in the context of individualized projects, 2) develop critical skills in big data analytics and machine learning-based methodology, and 3) answer key POR questions in RMDs and OA through a diversity and health equity lens.