PluDeeLea. Design, Implementation and Assessment of a Deep Learning Care Model for Multimorbid Patients to Aid Decision-making.

Call for tender for project: Grants for health research and development projects 2019: Promotion of health research

Funding body: Department of Health, Basque Government

Record no.: 2019222003

PI: Ane Fullaondo

Description: The project aims to develop and implement a Deep Learning system based on data collected from the medical history of multimorbid patients, in order to:

  • Classify multimorbid patients according to their care needs, so as to establish more personalised programmes.
  • Carry out early and automatic detection of patients who are going to become unstable and hence, be proactive with them.
  • This system can be implemented in day-to-day practice, integrated in a new healthcare model for multimorbid patients.


  • Design and develop a Deep Learning system based on the processing of Big Data from clinical data that enables the differentiation and classification of multimorbid patients, according to their healthcare needs.
  • Design, implement and assess a personalised medical model for multimorbid patients, based on systems that support decision-making.
  • Design patient healthcare model.
  • Implement the model in a pilot study.
  • Assess the implemented healthcare model.

Study design: The Deep Learning system is built based on all the data from multimorbid patients in the Basque Country who have been in the Osakidetza information system since 2011. The first population stratification was carried out that year and is updated periodically. Since then, these patients are marked with a multimorbid patient label, which is the requirement to be included in the project. The project will build the database by collecting anonymous information from Osakidetza’s OBI data extraction system. Once the information is filtered and analysed by experts, the final database will be defined, and it will be used to design predictive models.

Roles: Kronikgune is a partner in the project