I-Corps: Digital twin technology via synthetic data generation to predict outcomes of clinical trials

这个i-Corps项目的更广泛的影响/商业潜力是开发一种适用于实验数据收集昂贵或不可行的领域的平台技术。第一个应用将是临床试验领域，在这一领域，精准医学的趋势越来越大，但随机对照试验的试验往往极其昂贵和耗时。对于资源更为有限的中型制药公司来说，情况尤其如此。这项数字孪生技术为公司提供了合成数据，以帮助优化试验设计，并在如何分配有限的财力和人力资源方面控制风险。这项技术还可能在电子商务和精准农业等潜在应用领域具有深远的应用。I-Corps项目开发了一种新的因果推理技术，以患者为基础准确预测临床试验的结果。这项技术使用不同患者、治疗方法和疾病的历史临床数据，建立每个患者的数字孪生兄弟。这种方法的一个新好处是，尽管经过完整临床试验的患者数据非常稀缺，但该技术可以准确地模拟整个患者群体的结果。此外，这种方法提供了对这种数字双胞胎是如何创建的可解释性的，这是医疗应用中的一个关键特征。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势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 the development of a platform technology that is applicable to fields where experimental data collection is expensive or infeasible. A first application will be in the domain of clinical trials, where there is a growing trend towards precision medicine, yet experimentation with randomized control trials is often extremely costly and time consuming. This is particularly true for mid-sized pharmaceutical companies with more limited resources. The digital twin technology provides companies with synthetic data to help optimize trial design and control risks in how to allocate limited financial and personnel resources. The technology may also have far-reaching applications in potential application areas of e-commerce and precision agriculture.This I-Corps project develops a new causal inference technology to accurately predict outcomes of clinical trials on a patient-by-patient basis. This technology builds a digital twin of each patient using historical clinical data across different patients, treatments, and diseases. A novel benefit of this approach is that despite very scarce data on patients that go through the full clinical trial, the technology can accurately simulate the outcomes of the entire patient population. Additionally, this approach offers interpretability of how such digital twins are created, a critical feature in medical applications.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.