The workshop took place in a hybrid format and brought together a diverse group of stakeholders from research, healthcare, industry, public authorities, and international institutions. Discussions were framed by the broader theme “Trusted AI for Bharat & Czechia: Safety, Accountability and Human-Centric Innovation”, reflecting a shared commitment to advancing trustworthy AI approaches across regions and sectors.
Held in a hybrid format, the workshop brought together representatives from research, healthcare, public institutions and industry, creating a truly cross-disciplinary space for discussion. On behalf of the organising institutions, the event featured Prof. Vladimír Mařík (CIIRC CTU) and Vít Dočkal (INDRC), together with experts from key national and international bodies.
Among the participants were representatives of the Czech Metrology Institute, including Petr Braťka and Martin Koval, alongside representatives of the Ministry of Health of the Czech Republic and the Ministry of Industry and Trade of the Czech Republic. The healthcare perspective was enriched by participation from hospitals such as Oslo University Hospital, as well as applied research centres, startups, and international research organisations.
Discussion focused on concrete and pressing questions affecting the deployment of AI in medical practice today, including:
- How to assess whether and how specific AI systems fall under the Medical Device Regulation (MDR),
- How to approach risk classification in a manner that is both practical and legally defensible, and
- How to define validation metrics that reflect real-world trustworthiness, robustness and reliability in clinical use.
The international dimension of the dialogue was further strengthened by the presence of Dr. Prakash Shelat, First Secretary at the Embassy of India in Prague, whose participation underlined the relevance of the discussion beyond the European context.
Dr. Vít Dočkal, INDRC Director : “Our ambition is to establish a strong and lasting collaboration with partners from the Indian government, academia and industry. Trusted AI in healthcare is not a local challenge. It is a global one, and only through sustained international dialogue can we address it meaningfully.”
A central theme of the discussion was the challenge of translating regulatory frameworks into practical assessment methods for AI-based medical devices.
Rather than engaging in abstract theory, participants worked towards shared understanding and alignment across disciplines, acknowledging that regulation, safety and trust already play a decisive role in how AI systems are developed and evaluated.
Prof. Vladimír Mařík, scientific director of CIIRC CTU : “Without clear definitions of AI systems, their autonomy and intended use, it is impossible to assess safety or apply regulation in practice meaningfully. In medical devices, AI systems cannot be ‘proven safe’ in theory. Trust must be built through repeated testing and, for the foreseeable future, through human oversight between the system and the patient.”
This perspective resonated across stakeholder groups, highlighting the importance of iterative validation and responsible deployment in clinical settings.
From the perspective of public authorities, the workshop demonstrated a clear need for coordinated outputs that could support consistent interpretation of regulatory requirements, especially in the perspective of expected regulatory revisions at the EU level.
Dr. Petr Braťka, Head of Preclinical Evaluation Department, Project Leadership Manager, Product Reviewer Center for Medical Device Certification from the Czech Metrology Institute : “From my perspective, the most promising outcome is the clear convergence on the need to work towards a formalised conclusion, for example, in the form of a jointly agreed opinion. Such an output could serve as a valuable point of reference not only for developers and notified bodies, but also for other stakeholders, including public authorities, when navigating the practical application of MDR and AI-related requirements.”
By creating space for this type of interdisciplinary exchange, the workshop reinforced INDRC’s and CIIRC CTU’s role in bridging technology, clinical practice and regulation, including through joint initiatives such as the CLARA project. The organizers aim for the outcomes of this dialogue to contribute to a more consistent understanding among key stakeholders and to support collaboration not only within Czechia, but also across the EU and in international cooperation, particularly with India. This would mark an important step towards ensuring that AI systems used in healthcare are safe, accountable and human-centric.
Photos: CIIRC CTU/ Roman Sejkot, Tereza Pospíšilová