Assessing health AI
Read performance results, understand a confusion matrix and ask the right questions of a supplier.
Training and capability building
Training on artificial intelligence for health
QSTOM-IT designs training to help technical, business and leadership teams understand AI performance, limitations and the evidence they should request before using or funding a system.
Innovation and R&D leadersHealthTech and MedTech companiesData and AI teamsHealthcare professionals and business teams
Audiences
Who is it for?
Innovation and R&D leadersHealthTech and MedTech companiesData and AI teamsHealthcare professionals and business teamsInvestors and incubatorsSchools, universities and training organisations
Programme
Training topics
Data leakage and bias
Patient overlap, temporal leakage, sampling bias and cohort representativeness.
Robustness and generalisation
Subpopulations, drift, external validation, degraded data and error analysis.
Health AI metrics
Accuracy, sensitivity, specificity, ROC-AUC, PR-AUC, calibration, thresholds and operational impact.
Generative AI in health
Sources, hallucinations, evaluation, confidentiality and limits of use.
EU AI Act for technical teams
Understand the main regulatory themes and identify technical evidence to document without replacing formal regulatory advice.
Formats
Formats
- 1 to 2 hour awareness session
- Half-day workshop
- Full-day training
- Multi-session programme
- Practical case tailored to your context
In Lille, on site or remotely
Training can be delivered in Lille and Hauts-de-France, at your premises in France or remotely for European teams.
No fixed catalogue
The programme is built around participant level, data type, use cases and decisions. Examples can be adapted to your sector and tools.
FAQ
Training FAQ
Is EU AI Act training regulatory advice?+
No. It is designed to explain the framework, concepts and technical evidence. Formal legal or regulatory analysis should be performed by the appropriate professionals.
Can you train non-technical teams?+
Yes. Metrics and risks are explained through concrete cases without requiring advanced machine learning knowledge.
Can you deliver training in English?+
Yes. Training and materials can be delivered in French or English.
QSTOM-IT
Do you want your teams to challenge health AI more effectively?
Tell us who the participants are, the context and the decisions they need to make. We will build an appropriate format.