Ethical Perspectives Towards Using Smart Contracts for Patient Consent and Data Protection of Digital Phenotype Data in Machine Learning Environments

项目摘要 我们的母项目（NIH R 01 MH 125958）的特点是越来越多的研究收集高度重复的数据， 通用和敏感的数字表型分析（DP）数据，对推进科学和医疗保健具有巨大的前景。DP数据， 结合人工智能和机器学习（AI/ML），有望彻底改变临床应用， 在精神病学内外都有个性化医疗的研究。然而，DP数据也可能 患者的隐私和自决权面临风险，因为他们的披露能力越来越强， 预测-人类无法察觉的情绪和行为状态。现有的法律的保护并不充分 解决DP数据生态系统的新功能，使应用程序（包括货币化和前 DP数据的变更）与患者获益不直接相关，并可能使患者暴露于 很难预测。本提案的目的是确定实际、道德和技术方面的好处、挑战和挑战， 以及实施智能合约的激励措施：一种新兴的隐私设计技术， 增强患者对未来使用其DP数据的控制。我们的补充汇集了跨学科的前- 精通生物伦理学、医学人类学、决策科学/行为经济学和机器学习。LED 由定性和混合方法的专家，我们的团队将进行访谈，以确定不同的利益相关者 对潜在利益、挑战和伦理考虑的认识和技术理解， 使用智能合约来实现患者数据共享偏好，以收集用于精神病学重新评估的DP数据， 搜索（目标1）。根据目标1的研究结果，并使用行为经济学的见解，在目标2中，我们将建立一个模型， 最佳的“选择架构”，以激励不同利益相关者参与智能合约， 在道德上是合理的，符合利益相关者的利益。这一贡献意义重大，因为它将提供 知识对于支持在不断增长的DP数据生态系统中实现患者保护现代化的合作至关重要- 以复杂的社会、经济、技术和法律的关系为特点的项目。我们的方法是创新的， 它将DP数据生态系统视为由人与人之间的关系组成的民族志空间 参与者和技术同时出现，但缺乏以患者为中心的激励框架- 工作这项工作是可行的，因为我们的研究人员团队在道德，技术， 与DP数据使用相关的行为和政策问题，以及在大规模 合作项目，解决与新兴技术（包括AI/ML）相关的道德问题。

PROJECT SUMMARY Our parent project (NIH R01MH125958) is characteristic of a growing number of studies collecting highly gran- ular and sensitive digital phenotyping (DP) data with great promise to advance science and healthcare. DP data, in combination with artificial intelligence and machine learning (AI/ML) is poised to revolutionize clinical ap- proaches to personalized medicine both within and outside of psychiatry. However, DP data also poses potential risks to patients’ privacy and self-determination due to their growing capacity to reveal – and increasingly to predict – emotional and behavioral states undetected by humans. Existing legal protections do not adequately address novel features of the DP data ecosystem which enable applications (including monetization and ex- change of DP data) that are not directly linked to patient benefit and potentially expose patients to risks that are difficult to predict. The objective of this proposal is to identify practical, ethical and technical benefits, challenges and incentives for implementing smart contracts: an emerging privacy-by-design technology with promise to enhance patient control over future uses of their DP data. Our supplement brings together interdisciplinary ex- pertise in bioethics, medical anthropology, decision science/behavioral economics and machine learning . Led by an expert in qualitative and mixed methods, our team will conduct interviews to identify diverse stakeholders’ perceptions and technical understandings about potential benefits, challenges, and ethical considerations for using smart contracts to implement patient data sharing preferences for DP data collected for psychiatric re- search (Aim 1). Informed by Aim 1 findings and using behavioral economics insights, in Aim 2 we will model an optimal “choice architecture” to incentivize diverse stakeholders’ engagement with smart contracts in ways that are ethically justified and align with stakeholder interests. This contribution is significant because it will provide knowledge critical to support collaborations that modernize patient protections within a growing DP data ecosys- tem characterized by complex social, economic, technical and legal relations. Our approach is innovative in that it treats the DP data ecosystem as an ethnographic space comprised of relations between and among human actors as well as technologies emerging in parallel but in the absence of an incentivizing patient -centered frame- work. The work is feasible because our team of established investigators have expertise in ethical , technical, behavioral and policy issues related to DP data use and a track record of success working together on large collaborative projects addressing ethical issues related to emerging technologies including AI/ML.