MyIRE founder Mark Graves presented a case study of blockchain technology combined with easy to use tools and its impact on research integrity. We also demonstrated how this approach empowers doctors to participate in research as part of their daily practice.
Modules comprising the process of creating a hypothesis, testing it, and publishing
the results were composed into projects. Multiple projects were run through the
system. The results of one of the projects are ready to submit for peer review with
the expectation that the results will be published in a journal with a reference to
MyIRE. That project will be repeatable, in the manner described above, by anyone and
anywhere -- thereby minimizing instrumental bias. The project was forked to a
different node of the same specification. Alterations were made to individual data
collection elements within the forked study. Interfaces remained interoperable and
the project was auto-documenting. Underlying data was not transferred between
Internal validity and end-user acceptance of individual modules was tested via deployment across multiple organizations. We put into place legally binding contracts for decentralization of those modules with differential privacy restrictions that may be regulated. Module provenance and history were stored with the capacity for each organization to verify integrity. Modules were stored separately from data owned by each organization, so data integrity and module integrity can be verified independently.
Our four-node proof of concept infrastructure consisted of Network, Infrastructure, Compute, and Storage nodes. We used non-specialty, commercially available hardware with machines consuming 50% maximum capacity in terms of processing, RAM, and storage. Power optimization was not taken into account. Rather, the potential for maximal participation was targeted by optimizing for mass commercial availability. Control plane segregation, separation of concerns, and principles of least privilege were three factors also considered in our design.