MKB EFRO

Project objective

Ameretat Health B.V. carried out this feasibility study together with the eHealth Innovation Lab to assess the technical and economic feasibility of a cloud-based AI system for real-time polyp detection during colonoscopy.

The project responds to an important clinical need: reducing the number of missed polyps during colonoscopy and supporting earlier detection of colorectal cancer. The main objective was to determine whether the AI solution can perform fast and robustly enough for real-time clinical support, and whether there is a realistic route towards integration, market access, reimbursement and further clinical validation.

Activities carried out

• Technical feasibility was tested by building and evaluating a full processing pipeline for real-time AI support during colonoscopy.
• Robustness was explored using a collection of 13 open datasets from the colonoscopy domain and an experimentation pipeline for comparison with state-of-the-art models.
• Clinical usability and relevance were explored through interviews with 12 clinicians and 10 observations in a gastroenterology setting.
• Privacy, GDPR and Software as a Medical Device (SaMD) requirements were mapped through a gap analysis and regulatory preparation work.
• Market potential, competition, business models, reimbursement routes and the regulatory roadmap for MDR certification were investigated.

Project results

• A working processing pipeline was implemented and achieved an average processing speed of 10 ms per frame, meeting the real-time clinical support requirement of at least 60 frames per second with low latency.
• The study confirmed that the solution has a strong technical basis for local deployment on a machine with a graphics card, with optional next steps towards edge or cloud computing.
• Clinician input confirmed opportunities not only in polyp detection, but also in areas such as cancer classification, reporting, planning and training.
• The compliance work showed that GDPR and SaMD requirements are technically feasible to address, and a high-level MDR/SaMD roadmap was prepared for the next development phase.
• The economic feasibility was assessed positively. The AI endoscopy market is growing strongly, and Ameretat’s cloud-native, software-only approach with CADx potential, real-time sizing and a pay-per-measurement model provides a differentiated position.
• The overall feasibility assessment was positive. The recommended next step is to continue with problem-solution fit validation, MVP development, clinical pilot preparation and the formal MDR certification pathway.

Financial support

Ameretat Health B.V. received a subsidy amount of €45,000 for this project. The financial support from the European Union enables Ameretat to accelerate the development and validation of medical AI technology for real-time decision support in endoscopy.

Contribution to Northern Netherlands

The project contributes to the Regional Innovation Strategy (RIS3) of Northern Netherlands, especially in Health & Care and Digitalisation. By supporting higher detection rates and earlier detection of colorectal cancer, the solution can contribute to better patient outcomes and more cost-effective healthcare. The collaboration with the eHealth Innovation Lab and regional medical experts also strengthens regional knowledge development in MedTech and advanced AI applications for healthcare.

Co-financed by the European Union, EZ and SNN

The logos are shown with the project description to make the financial support from the European Union, EZ and SNN visible.