Currently submitted to: JMIR AI
Date Submitted: Jun 7, 2026
Open Peer Review Period: Jun 12, 2026 - Aug 7, 2026
(currently open for review)
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
A Human–AI Collaboration Framework for Rare Disease Clinical Trials: Methodological Synthesis from the REiNS AI Consortium in Neurofibromatosis and Schwannomatosis
ABSTRACT
Background:
The REiNS AI Consortium identified a critical gap in rare disease clinical trials: no principled framework exists to guide decisions about appropriate automation levels, validation requirements, and oversight structures for AI applications in data-limited settings.
Objective:
To address this, the Consortium developed a Human–AI Collaboration Framework synthesized across four domains (Data, Accessibility, Diagnostics, and Histology) using neurofibromatosis (NF) and schwannomatosis (SWN) as the development and validation context.
Methods:
Expert consensus was developed through structured presentations and moderated discussions convened by the REiNS AI Consortium. Recurring tensions between automation and oversight were synthesized into a two-axis framework defined by automation level and validation requirement. The framework was refined through cross-domain comparison and applied across each AI application discussed by the consortium.
Results:
Across domains, AI applications most consistently clustered at the AI-Augmented / Research-Grade interface, where automation improves efficiency while expert oversight preserves interpretability and reliability. A recurrent deployment risk emerged when automation level and evidentiary requirements were poorly aligned. Domain-specific instantiations include FAIR-aligned curation pipelines; dual-interface LLM architectures for patient communication; validated deep learning segmentation models for cutaneous neurofibromas, plexiform neurofibromas, and vestibular schwannomas; and computational pathology frameworks with prospective NF and SWN application pathways. Patient representatives identified AI fairness, transparency, and human-centered care as non-negotiable requirements across all domains.
Conclusions:
The Human–AI Collaboration Framework provides a transferable methodology for rare disease and oncology trials, demonstrating that multi-stakeholder expert synthesis is a viable mechanism for generating AI governance frameworks in data-limited settings.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.