Periodic Reporting for period 2 - StrokeFlow (Empowering physicians to provide every stroke treatment with the right treatment in time, with StrokeFlow)
Berichtszeitraum: 2024-03-01 bis 2025-02-28
Stroke remains one of society's most significant healthcare challenges, representing the leading cause of severe long-term disability and carrying an immense societal burden, with costs exceeding €80 billion annually in the EU alone. The disease is exceptionally time-critical; for every minute treatment is delayed, patients lose an average of 22 healthy life-days. Despite promising new therapies, a critical gap has persisted in stroke care: the inability for specialists to rapidly obtain the precise, patient-specific information needed to select the optimal treatment in the hyper-acute phase. This has resulted in a largely "one-size-fits-all" approach, limiting the potential for improved patient outcomes.
The overall objective of the StrokeFlow project was to address this critical need by developing and validating a paradigm-shifting, AI-powered software platform. The project aimed to empower physicians to make true, personalized treatment decisions by delivering revolutionary technology that provides unprecedented insights from standard, readily available medical scans. The core goals were to drastically reduce the time to intervention, enhance the quality of diagnostic information, and create a scalable, secure, and clinically integrated solution ready for widespread clinical adoption.
Over the course of the project, we successfully developed a near-market, regulatory-ready solution that directly tackles the identified challenges. The project delivered novel AI methodologies capable of deriving advanced diagnostic information from routine scans, a major step toward democratizing access to high-quality stroke care. Furthermore, a scalable, cloud-agnostic platform was built and validated, demonstrating a clear pathway for seamless integration into existing hospital workflows. By achieving these objectives, the project has laid a robust foundation to transform stroke care from a standardized process into a data-driven, personalized discipline.
The project's results are expected to generate impacts of significant scale.
The overall objective of the StrokeFlow project was to address this critical need by developing and validating a paradigm-shifting, AI-powered software platform. The project aimed to empower physicians to make true, personalized treatment decisions by delivering revolutionary technology that provides unprecedented insights from standard, readily available medical scans. The core goals were to drastically reduce the time to intervention, enhance the quality of diagnostic information, and create a scalable, secure, and clinically integrated solution ready for widespread clinical adoption.
Over the course of the project, we successfully developed a near-market, regulatory-ready solution that directly tackles the identified challenges. The project delivered novel AI methodologies capable of deriving advanced diagnostic information from routine scans, a major step toward democratizing access to high-quality stroke care. Furthermore, a scalable, cloud-agnostic platform was built and validated, demonstrating a clear pathway for seamless integration into existing hospital workflows. By achieving these objectives, the project has laid a robust foundation to transform stroke care from a standardized process into a data-driven, personalized discipline.
The project's results are expected to generate impacts of significant scale.
Throughout the project, our activities focused on the core scientific and technical development required to build the StrokeFlow platform. The main achievements are the result of extensive research, engineering, and validation work.
Development of Novel AI for Clot Analysis: A primary focus was creating technology to automatically identify and analyze blood clots from standard CT scans. We successfully developed a foundational "digital biopsy" algorithm and, through rigorous testing, gained critical insights into its performance. This work established a validated methodology and a clear path for the continued refinement required for a final, regulatory-approved product.
Breakthrough in CTP-Free Perfusion Imaging: A key scientific achievement was the development of an AI model that can accurately estimate brain perfusion—a critical factor in stroke diagnosis—using only routine scans. This eliminates the need for more advanced, time-consuming, and less accessible imaging techniques, representing a major step toward democratizing high-quality stroke care.
Successful Integration into the Clinical Environment: We achieved a crucial technical milestone by seamlessly integrating our AI-generated visualizations directly into the surgical environment (the angio suite). This allows physicians to see and interact with the AI results in real-time during treatment, a major step in translating our technology into tangible clinical value.
Creation of a Scalable and Secure Platform: Significant engineering efforts were dedicated to building a robust, secure, and highly reliable cloud architecture. This modern infrastructure ensures that StrokeFlow can be deployed efficiently across international hospitals, guaranteeing the high performance and stability required in a critical care setting.
User-Centric Design and Validation: In close collaboration with clinicians, we developed and implemented a complete and intuitive user experience (UX) framework. Through multiple rounds of user testing and feedback, we ensured the platform is not only technologically advanced but also practical and easy to use in the high-pressure environment of acute stroke care.
Development of Novel AI for Clot Analysis: A primary focus was creating technology to automatically identify and analyze blood clots from standard CT scans. We successfully developed a foundational "digital biopsy" algorithm and, through rigorous testing, gained critical insights into its performance. This work established a validated methodology and a clear path for the continued refinement required for a final, regulatory-approved product.
Breakthrough in CTP-Free Perfusion Imaging: A key scientific achievement was the development of an AI model that can accurately estimate brain perfusion—a critical factor in stroke diagnosis—using only routine scans. This eliminates the need for more advanced, time-consuming, and less accessible imaging techniques, representing a major step toward democratizing high-quality stroke care.
Successful Integration into the Clinical Environment: We achieved a crucial technical milestone by seamlessly integrating our AI-generated visualizations directly into the surgical environment (the angio suite). This allows physicians to see and interact with the AI results in real-time during treatment, a major step in translating our technology into tangible clinical value.
Creation of a Scalable and Secure Platform: Significant engineering efforts were dedicated to building a robust, secure, and highly reliable cloud architecture. This modern infrastructure ensures that StrokeFlow can be deployed efficiently across international hospitals, guaranteeing the high performance and stability required in a critical care setting.
User-Centric Design and Validation: In close collaboration with clinicians, we developed and implemented a complete and intuitive user experience (UX) framework. Through multiple rounds of user testing and feedback, we ensured the platform is not only technologically advanced but also practical and easy to use in the high-pressure environment of acute stroke care.
A validated framework for 'digital clot biopsy'
The current state of the art leaves physicians to treat blood clots with very limited information on their specific characteristics. Our technology will provide deep physiological insights directly from standard scans in the hyper-acute setting. This result creates a pathway to move away from a "one-size-fits-all" treatment protocol towards true, personalized intervention, where the treatment is tailored to the unique nature of the clot.
The ability to generate advanced perfusion data from routine imaging
Previously, obtaining critical data on brain tissue health required specialized, time-consuming, and costly CTP or MRI scans, which are unavailable in many hospitals. Our project delivered an AI-based solution that generates this essential information from the routine scans every stroke patient already receives. This breakthrough democratizes access to advanced stroke diagnostics, making high-quality assessment more equitable and immediately available everywhere.
A seamless integration of AI insights into the live interventional environment
Typically, AI analysis exists in a separate diagnostic environment, with results communicated to the interventionist. Our work has successfully embedded dynamic, interactive AI visualizations directly into the angio suite used during surgery. This brings actionable intelligence to the treating physician in real-time at the point of care, a paradigm shift that enhances procedural efficiency and decision-making.
These innovations have established a new technological benchmark. The key need to ensure their successful uptake is now to bridge from technical validation to widespread clinical adoption. This involves completing the final phase of regulatory validation and executing a focused market access strategy to ensure these state-of-the-art results can begin delivering patient impact on a global scale.
The current state of the art leaves physicians to treat blood clots with very limited information on their specific characteristics. Our technology will provide deep physiological insights directly from standard scans in the hyper-acute setting. This result creates a pathway to move away from a "one-size-fits-all" treatment protocol towards true, personalized intervention, where the treatment is tailored to the unique nature of the clot.
The ability to generate advanced perfusion data from routine imaging
Previously, obtaining critical data on brain tissue health required specialized, time-consuming, and costly CTP or MRI scans, which are unavailable in many hospitals. Our project delivered an AI-based solution that generates this essential information from the routine scans every stroke patient already receives. This breakthrough democratizes access to advanced stroke diagnostics, making high-quality assessment more equitable and immediately available everywhere.
A seamless integration of AI insights into the live interventional environment
Typically, AI analysis exists in a separate diagnostic environment, with results communicated to the interventionist. Our work has successfully embedded dynamic, interactive AI visualizations directly into the angio suite used during surgery. This brings actionable intelligence to the treating physician in real-time at the point of care, a paradigm shift that enhances procedural efficiency and decision-making.
These innovations have established a new technological benchmark. The key need to ensure their successful uptake is now to bridge from technical validation to widespread clinical adoption. This involves completing the final phase of regulatory validation and executing a focused market access strategy to ensure these state-of-the-art results can begin delivering patient impact on a global scale.