Exploring Stakeholder Requirements to Enable Research and Development of Artificial Intelligence Algorithms in a Hospital-Based Generic Infrastructure: Results of a Multistep Mixed Methods Study
Bibliographic record
Abstract
BACKGROUND: Legal, controlled, and regulated access to high-quality data from academic hospitals currently poses a barrier to the development and testing of new artificial intelligence (AI) algorithms. To overcome this barrier, the German Federal Ministry of Health supports the "pAItient" (Protected Artificial Intelligence Innovation Environment for Patient Oriented Digital Health Solutions for developing, testing and evidence-based evaluation of clinical value) project, with the goal to establish an AI Innovation Environment at the Heidelberg University Hospital, Germany. It is designed as a proof-of-concept extension to the preexisting Medical Data Integration Center. OBJECTIVE: The first part of the pAItient project aims to explore stakeholders' requirements for developing AI in partnership with an academic hospital and granting AI experts access to anonymized personal health data. METHODS: We designed a multistep mixed methods approach. First, researchers and employees from stakeholder organizations were invited to participate in semistructured interviews. In the following step, questionnaires were developed based on the participants' answers and distributed among the stakeholders' organizations. In addition, patients and physicians were interviewed. RESULTS: The identified requirements covered a wide range and were conflicting sometimes. Relevant patient requirements included adequate provision of necessary information for data use, clear medical objective of the research and development activities, trustworthiness of the organization collecting the patient data, and data should not be reidentifiable. Requirements of AI researchers and developers encompassed contact with clinical users, an acceptable user interface (UI) for shared data platforms, stable connection to the planned infrastructure, relevant use cases, and assistance in dealing with data privacy regulations. In a next step, a requirements model was developed, which depicts the identified requirements in different layers. This developed model will be used to communicate stakeholder requirements within the pAItient project consortium. CONCLUSIONS: The study led to the identification of necessary requirements for the development, testing, and validation of AI applications within a hospital-based generic infrastructure. A requirements model was developed, which will inform the next steps in the development of an AI innovation environment at our institution. Results from our study replicate previous findings from other contexts and will add to the emerging discussion on the use of routine medical data for the development of AI applications. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.2196/42208.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.155 | 0.132 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.004 | 0.003 |
| Science and technology studies | 0.005 | 0.004 |
| Scholarly communication | 0.006 | 0.005 |
| Open science | 0.003 | 0.009 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".