The Gentil – Text Mining project won the Portugal Digital Awards in the Best Future of Work Project category, which acknowledges the best initiative related to employees’ productivity and effectiveness. Developed by INESC TEC in partnership with the Portuguese Institute of Oncology of Porto – Francisco Gentil (IPO Porto) and ValeConsultores, the technology uses text-mining and Natural Language Processing (NLP) tools to provide structured clinical data, supporting decision-making in oncological contexts.
The main objective of the team of researchers is to facilitate the access and consultation of clinical data to support healthcare professionals in their daily activities. “The Gentil – Text Mining project enables the structured extraction of data from clinical records containing information about the patient’s health status, pathological history, biometric data, exams and diagnoses, among others, making it available on easy-access panels for healthcare professionals”, explained Mário Amorim Lopes, researcher at INESC TEC.
In order to process the text written by the physicians in the clinical records, the team resorted to NLP techniques. Alípio Jorge, INESC TEC researcher and head of the project’s NLP team, stated that “it is necessary to apply computational techniques for natural language processing, which help the machine to understand the content, while identifying diagnoses, treatments and future actions, among other elements, in the texts. This is done using a set of techniques including neuronal language models and Entity Linking”. Moreover, there’re other challenges, since the texts are written in Portuguese, a language that is less studied among the scientific community.
In this sense, the project presents a solution for the large amount of data associated with each patient and, consequently, for the time it takes to consult said data, which is often generated in institutions or by different healthcare professionals.
By saving time in the consultation process, it will be possible to eliminate redundancies, i.e., duplicated exams, and to reduce the time needed to treat patients. “The tool will also make it possible to provide statistical information/clinical evidence from patients part of similar cohorts, allowing the effectiveness of certain treatments to be compared, as well as cross-referencing information with American and European treatment protocols, as well as scientific articles”, mentioned researcher Mário Amorim Lopes.
The technology will mainly serve IPO-Porto’s healthcare professionals, towards reducing consultation, diagnosis, and treatment times for patients in this unit, as well as the associated costs. “With Gentil – Text Mining, the physicians will no longer need to consult pages and pages of ‘free text’, and will now have access to a dashboard that includes critical information to support the decision-making process”, emphasised Renato Magalhães, project manager and Director of the Information and Communication Systems Management Service at IPO Porto.
Concerning the acknowledgement at the Portugal Digital Awards 2021, the team stated that “it awards the work developed and provides additional motivation to keep improving it”. The Portugal Digital Awards are promoted by IDC and Axians; the sixth edition acknowledged initiatives that promote digital transformation, in 22 different categories.
The INESC TEC researchers mentioned in this news piece are associated with UP-FCUP and UP-FEUP.
News, current topics, curiosities and so much more about INESC TEC and its community!
