This program develops decision support tools that clinicians can inspect, question and improve, with a focus on transparency, robustness and responsible use in practice.
The use of AI in diagnostics and triage continues to expand, but adoption depends on more than performance metrics. Tools that affect health-related decisions must be understandable, governed and tested in settings where uncertainty, time pressure and accountability are real.
The Trustworthy Clinical AI program studies interpretable models, fairness assessment, data provenance, human oversight and integration into care workflows. We are particularly interested in how decision support can strengthen professional judgment rather than obscure it.
Areas of focus
Transparent model logic and output explanations
Validation across multiple environments
Bias and fairness testing Human oversight and escalation design
Governance, documentation and accountability
Meridian's Clinical AI packages - for all needs
Meridian One
Our entry level with enhanced testing for clinical environments
€100/mo
Core AI model
1000 Queries
Validation for 1 environment
Meridian Two
Our mid-tier suitable for most scenarios
$250/mo
Core AI model
Evolution AI addon
10 000 Queries
Validation for 10 environments
Testing tools
Meridian Full
For large scale high end scenarios
Price by quote
Full features from Meridian One and Two, infinite scenarios and full suite of current development tools.
Why trust matters
Trust is built when users understand what a system is intended to do, what its limitations are and how outputs should be interpreted. For this reason, the program combines model development with documentation practices, evaluation criteria and interface design approaches that support critical use.