#SeminarIRSJD · The Use of Deep Learning for Patient Stratification



If you wish to obtain the connection link, contact with frecerca.comunicacio(ELIMINAR)@sjd.es (exclusive seminar for staff of the institution).


Medical data are collected by healthcare systems all over the world in a daily basis, including diagnosis evidence and prescriptions among others. To convert this large amount of data into knowledge remains challenging and an open issue.

In this talk showcase how we use AI specially tuned for clinical applications. We will first explain how to convert self-declared symptoms into ontology annotations for deep phenotyping-based patient stratification. Secondly, we will portrait how to use deep autoencoders to observe the longitudinal stratification in a DM2 large population from primary care data.