I-Corps: diagnostic & patient management tool for physicians treating movement disorders

这个I-Corps项目更广泛的影响/商业潜力是加速神经精神化合物的药物开发过程，并改善脑部疾病的临床管理。该项目将评估基于深度学习的姿势跟踪算法的进步是否可以在生物技术/制药和医疗保健领域带来新的商机。行为是大脑过程复杂协调的结果，并受到疾病或损伤的影响。患者或动物的行为库是关于其神经精神健康的丰富的非侵入性数据来源，但存在有限的工具来挖掘这种类型的数据。解决行为量化的问题有可能使神经精神药物的开发更有效，因为生物技术/制药公司可以更好地了解他们的药物从首次在动物身上测试到关键的人体研究是如何工作的。行为量化工具还可以帮助临床医生治疗神经精神疾病，由于其复杂性，通常需要广泛的培训才能正确诊断和管理。 这个I-Corps项目是基于开发一种基于深度学习的行为量化工具，用于药物开发和临床医学。目前的方法来衡量神经精神干预的疗效是有限的，其灵敏度，效率和可重复性。神经科学的突破性进展为药物开发者阐明了新的靶向机制，但评估新疗法疗效的技术有限。此外，用于神经精神测试的动物模型对临床结果的可转化性有限，使得药物开发人员难以进行资产优先级排序。基于深度学习的行为测量工具通过自动化具有高灵敏度和特异性的行为分析来作为这个问题的非侵入性解决方案。这些工具可以提取药物开发人员目前无法获得的见解，这不仅会促进对疾病生物学的理解，而且还会收集药物批准的关键数据。生物技术/制药公司能够在所有开发阶段对治疗效果进行强有力的表征，这降低了目前与神经精神药物开发相关的监管和商业风险。同样，医生将能够使用先进的行为量化技术，将患者的症状与“数字指纹”库进行比较，以准确诊断病情和管理症状。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact/commercial potential of this I-Corps project is to accelerate the drug development process for neuropsychiatric compounds and improve the clinical management of brain disorders. This project will evaluate whether advances in deep learning-based pose tracking algorithms can lead to new business opportunities within both biotech/pharma and healthcare. Behavior is a result of the complex coordination of brain processes and is affected by disease or impairment. The behavioral repertoire of a patient or animal is a rich noninvasive source of data about its neuropsychiatric health, but limited tools exist to mine this type of data. Solving the problem of behavioral quantification has the potential to make neuropsychiatric drug development more efficient by providing biotech/pharma with a better understanding of how their drugs are working from the time they are first tested in animals all the way through pivotal human studies. Behavioral quantification tools can also assist clinicians with the treatment of neuropsychiatric diseases, which due to their complexity typically require extensive training to properly diagnose and manage. This I-Corps project is based on the development of a deep learning based behavioral quantification tool for use in drug development and clinical medicine. Current approaches to measure the efficacy of neuropsychiatric interventions are limited by their sensitivity, efficiency, and reproducibility. Breakthroughs in neuroscience have elucidated new mechanisms for drug developers to target, but limited techniques exist to evaluate the efficacy of new treatments. In addition, animal models for neuropsychiatric testing have limited translatability to clinical results, making asset prioritization difficult for drug developers. Deep learning based behavioral measurement tools serve as a noninvasive solution to this problem by automating behavioral analysis with high sensitivity and specificity. These tools can extract insights that are currently inaccessible to drug developers, which will not only lead to advances in the understanding of the biology of disease, but also the collection of pivotal data for drug approval. The ability of biotech/pharma to robustly characterize therapeutic efficacy across all stages of development reduces the regulatory and commercial risks currently associated with neuropsychiatric drug development. Similarly, physicians would be able to use advanced behavioral quantification techniques to compare a patient’s symptoms against a library of “digital fingerprints” to accurately diagnose conditions and manage symptoms.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.