Big Data method to predict covid patient outcomes with 90% accuracy

The Lucentia research group, from the University of Alicante, is developing the project ‘Big Data and Artificial Intelligence Solution to improve the diagnosis and medical care of those affected by COVID-19’, in collaboration with several hospitals in the Valencian Community.

 

Through the project, the aim is to predict the evolution of infected patients so that hospitals know in advance whether it will be a mild case, require hospitalization or present a high risk of death. It is an application that extracts information related to patients’ previous pathologies, results and tests to design personalized treatments and the evolution of the disease. Based on the concurrence of these data, the algorithms suggest the most successful treatment for each patient.

 

The Councilor for Finance and Economic Model, Vicent Soler, and the Councilor for Innovation, Carolina Pascual have presented today with the rector of the University of Alicante, Amparo Navarro, the initiative that has been funded through donations made by citizens, through the line launched by the Generalitat. Specifically, this project received funding of 100,000 euros.

 

This is one of the projects chosen by the Regional Ministry of Innovation to allocate part of the 1.2 million received from COVID-19 donations. “We are aware that sustained investment in the field of science allows us to get out of the crisis faster and better, so we have not hesitated in which areas we had to prioritize.

 

It is a project with a model that offers 90% accuracy and developed in an open format so that it can be easily exportable to all hospitals in the State.

 

The proposal and construction of this model is based on the previous experience of the research group Lucentia of the University of Alicante.
The project has allowed the development of a platform that accesses patients’ medical records to extract information on previous pathologies, treatments received and tests performed to detect COVID-19.

 

Based on this data, artificial intelligence models are designed and trained so that, each time a new patient is admitted, the system classifies them into one of the three groups.
In addition, according to data on the evolution of all the patients to the measures and treatments received, the algorithms will suggest the most successful treatment for each one of them.

 

 

 

 

 

More informations:

 

Levante (EMV) – Un método de big data predice la evolución de los pacientes covid con una precisión del 90 %

 

Cadena SERLa Inteligencia Artificial puesta al servicio de una mejor atención al paciente de COVID-19

 

 

Source: University of Alicante