Partners in the iToBoS consortium recently had their work published in a peer reviewed scientific journal, entitled Frontiers in Digital Health.
The article was published as part of a special issues, entitled Trustworthy AI in healthcare. The journal is open-access and seeks to publish noteworthy contributions at the intersection of health and digitalisation.
The article was a joint effort by Trilateral Research and IBM, who have been working together on WP2 - Privacy, data protection, ethical and societal issues in iToBoS solutions, and WP4 - Implementation of AI privacy and anonymization throughout the project.
The article, entitled “Dicing with data: the risks, benefits, tensions and tech of health data in the iToBoS project”, provides an introduction to certain AI, data, privacy and explainability aspects of the project - including technologies, risk assessment methods, and an introduction to how the consortium have attempted to mitigate specific risks within the project.
The paper outlines four of the more complex aspects of the project:
- the relatively low population clinical trial study cohort, which poses risks associated with data distinguishability and the masking ability of the applied anonymisation tools.
- the project's ability to obtain informed consent from the study cohort given the complexity of the technologies.
- the project's commitment to an open research data strategy and the additional privacy risk mitigations required to protect the multi-modal study data.
- the ability of the project to adequately explain the outputs of the algorithmic components to a broad range of stakeholders.
The paper also discusses how the Melanoma Patients Network Europe (MPNE), a partner in the iToBoS project, have been key figures in trying to understand the implications of these issues, and how they relate to wider tensions in the health sector.
MPNE have provided an important avenue for discussion and debate about the wider qualification of risks associated with health data and provided an avenue for gathering the perspective of patients and their advocates about the benefit/risk trade-off that they face, which has helped the experts involved in the iToBoS project better understand the complex interplay of privacy, explainability, and AI in the health sector.
Find out more about the journal at Frontiers in Digital Health.