Using artificial intelligence to improve stroke patient outcomes
Using artificial intelligence to improve stroke patient outcomes
As part of Brain Week, Stroke Alliance for Europe (SAFE) presents an interview with Dietmar Frey from the German Charité Universitätsmedizin Berlin, one of the partners of the VALIDATE project, about this new and innovative project.
SAFE: What is VALIDATE?
Dietmar Frey:
VALIDATE is an EU-funded research project to develop and validate an AI-based prognostic tool for predicting health outcomes in acute ischemic stroke patients. A decision support system will provide physicians and healthcare professionals with additional information to select optimal treatments to improve health outcomes and quality of life.
What is different and new about this research?
Dietmar:
By leveraging available medical data, leveraging AI technology capabilities, and developing and validating trusted AI solutions to be implemented in clinical practice, we aim to advance the current state of clinical science by making a significant and lasting impact on acute stroke care, improving patient survival and outcomes. We are the first to integrate machine learning and deep learning models into of a comprehensive decision support system in the area of stroke.
What will this new technology mean for stroke patients?
Dietmar:
One direct impact on the stroke patient will be that treatment decisions will be supported by data-driven analytics. In other words, treating physicians will have all the information available to make the best treatment decisions for a particular patient. Our results should be the basis for the inclusion of prognostic and predictive approaches based on artificial intelligence in clinical guidelines for the treatment of acute stroke. Our approach will ensure that biases are taken into account, that subgroups of patients are not discriminated against, and that transparency and safety are paramount. This requires testing with patients and healthcare professionals, including validation of user interface, user experience, as well as validation of predictive explanations (AI). Because patients must be at the center of development, we involve patient organizations and patients in the development and testing process. In addition, patients and their families must be given sufficient information about their disease and what decisions have been made using the AI-based prognostic tool through tested communication flows.
When will this technology start being used in clinical practice?
Dietmar:
Translating our findings into clinical practice will have several sequential steps. As part of this project, we will clinically validate the demonstration system in a clinical setting by conducting a multicenter prospective study. We will develop a plan for the regulatory acceptability of a validated forecasting tool. The medical benefit of the tool must be proven before it is introduced into the work process. This will be done – after the successful completion of the study within the VALIDATE time schedule – after the end of the project. One of the main outputs would be the preparation for a randomized controlled clinical trial (RCT) as a next step.
Will this technology save lives?
Dietmar:
This technology has the potential to save lives and improve outcomes for stroke patients by matching each patient to the best available treatment. By leveraging the power of data and derived models, the integration of individual patient characteristics will lead to more individualized approaches and may have a major impact on survival and treatment outcomes in acute stroke.
VALIDATE received funding from the European Union’s Horizon 2020 research and innovation program under Grant Agreement No. 777107.
More information can be found at https://validate-project.eu/
