Ai will help neurologist manage stroke and predict prognosis

by time news

2023-06-20 15:16:47

Artificial intelligence will help the neurologist manage stroke and predict its prognosis. This is the aim of the TruStroke project, with which the Gemelli Irccs University Hospital Foundation was the winner of the prestigious Horizon tender of the European Commission, together with 12 other partners. The project, focused on the use of artificial intelligence in stroke patients, will be funded for a total amount of 6 million euros. The role of Gemelli will be to coordinate the clinical work package (Wp6) of the project.

To arrive at the development of a new platform based precisely on artificial intelligence, at the service of doctors, patients and caregivers, it is necessary to ‘educate’ the artificial intelligence algorithms to make them learn how stroke evolves in a large group of patients. Once the preparatory phase has been completed, these AI tools will be validated on a series of patients, to verify if they have been correctly ‘trained’, i.e. to evaluate whether the prediction that the algorithm makes in terms of clinical improvement or worsening, of stroke recurrence , of re-hospitalization, of adherence to secondary prevention therapy, is reliable or not.

“Artificial intelligence – explains Pietro Caliandro, medical director of the Stroke Unit of the Gemelli Polyclinic, professor of Neurology at the Catholic University, Rome campus, and coordinator of the project’s clinical Work Package – has enormous computing power and could reveal to us, in a very early stage, a series of information that currently escapes us. days or 3, 6, 12 months, in terms of survival and disability development. and hospitalization times, but also at home, where it will continue to be followed through tools and applications that will act as an interface between the patient at home and us in the hospital”.

The TruStroke project is divided into two large strands: the technical part, which will develop the algorithms, and the clinical part, coordinated by the Gemelli Polyclinic. The technical partners are Eurecat, Josef Stefan Institute, Nora Health, Politecnico di Milano, Cnr, Nacar. A federated learning infrastructure (hosted by CERN) will also be developed and used to enable collaborative creation of reliable AI-based predictive models, leveraging stroke data, without compromising privacy (Eatris, Stroke Alliance for Europe) and implementing the best security protocols.

The Gemelli Irccs Polyclinic Foundation and the Gemelli Generator multidisciplinary center, directed by Professor Vincenzo Valentini, will have the role of coordinating the clinical activities carried out in hospital facilities of global importance (Vall D’Hebron University Hospital in Barcelona, ​​Catholic University of Louvain, University of Ljubljana); they will define the clinical variables on which the predictive models will be based, promote clinical studies to verify the effectiveness of the developed predictive models and validate them clinically; they will also contribute to the integration of the results into international guidelines that define the salient aspects of the care of stroke patients.

“The TruStroke project – explains Paolo Calabresi, director of the Complex Operational Unit of Neurology at the Gemelli Polyclinic and full professor of Neurology at the Catholic University – represents an extraordinary opportunity to implement the management processes of patients with ischemic stroke with important repercussions for the individual patient and the community”. The general coordination of the network is entrusted to the University Hospital of Vall D’Hebron in Barcelona.

#neurologist #manage #stroke #predict #prognosis

You may also like

Leave a Comment