Overview
The Precision Psychiatry Wiki is a source-based partner discovery and project formation resource. It explains how foundational research and AI capability at the Neural Information Processing Group of Eötvös Loránd University connect with clinical questions and data collection at Semmelweis University and other sites, speech and language technology at the University of Szeged, product development at Argus Cognitive, large-scale Danish research environments, and industrial translation.
The aim
Build an ecosystem connecting basic research, well-specified data collection, reproducible analysis, clinical validation, and protectable or deployable diagnostic, monitoring and therapeutic support methods.
Throughout, clinical validity, clinical utility, intellectual property, regulatory status, and commercial readiness are separate dimensions and are described separately. → Capability readiness
What the wiki supports
- Presenting existing research, technical, clinical and translational capabilities.
- Identifying data resources and opportunities for harmonised new data collection.
- Converting collaboration ideas into reviewable and auditable projects.
- Matching projects with national, European, foundation and industrial funding.
The ecosystem at a glance
| Tier | Who | What they bring |
|---|---|---|
| 1 — Hub | NIPG / ELTE | The method layer every clinical use case needs |
| 2 | Semmelweis / VIKOTE | Funded avatar-therapy trial, real patient population |
| 2 | Speech Gap | The patented measurement technology being integrated |
| 2 | Argus Cognitive | The clinical-AI translation case study |
| 3 | Michael E. Benros | Danish registers, biobanks, external validation potential |
| 3 | Erik Szabó | Capital, incubation, and the viability screen |
| 4 | Supporting collaborations | Sorbonne, Delft/MeMo, DFKI, Richter — credibility and transferable experience |
Full detail with firmness ratings at Ecosystem map.
The concrete near-term project
Speech Gap integrated into the funded VSMT avatar-therapy project as an exploratory, non-comparative longitudinal study: a 12-week therapy period with weekly short speech measurements, and three or four in-depth BPD reference assessments.
The analyses concern within-person trends and prediction of later reference scores. A separate untreated control group is not intrinsic and is not needed for that observational question. Planned safeguards are ClinicalTrials.gov registration and a PCI Registered Report — registration, ethics approval and Stage 1 in-principle acceptance being three separate, not-yet-completed processes.
The clearest technical contribution NIPG would make to it is the respiratory signal extension: the current software treats respiratory pauses as silence, which may discard clinically relevant information in BPD.
The convening event
The INNOFAIR 2026 workshop — Precision Psychiatry, Translated: Speech, Avatars, and the Path to Clinical AI — brings six speakers and a closing roundtable together to produce a concrete map of collaboration opportunities across this ecosystem.
How to navigate
- New here? Start with NIPG, then Speech Gap, then the workshop programme.
- Assessing a collaboration? Ecosystem map → Capability portfolio → Collaboration process.
- Checking what is actually claimed? Editorial safeguards and Capability readiness.
- Everything: Index.
Governing principle
Every page in this wiki respects Editorial safeguards — hard constraints on inventorship, institutional relationships, and the limits of what current evidence supports. Where a source hedges, the hedge stays.