The UMCG website shows YouTube videos. YouTube places cookies if you watch them. Choose ‘Rather not’ if you do not want these cookies. We also place cookies of our own. These help us improve the usability of our website.
More on our cookie policy.
Led by epidemiologist Dr Lilian Peters, the research team uses a smart language processing model to extract valuable information from doctors notes in the digital patient records of Dutch GPs. The team, consisting of experts from UMCG, Radboud UMC, Maastricht UMC and Erasmus MC, proved that this method can detect COVID-19 infections very accurately. It also proved to work in the early stages of the pandemic when testing was still limited and even before the first Covid-19 patient in the Netherlands was officially announced.
The study used data from the AHON-database and brought together knowledge from general practitioners, epidemiologists, microbiologists and data scientists. The developed model predicted the COVID-19 outbreak a few weeks before the first patient was hospitalised in the Netherlands. Maarten Homburg, general practitioner and PhD student within the project, explains: "We used BERT, a Dutch language model, to detect COVID-19 in GP data. By training the language model with symptom information, it learned to recognise COVID-19 based on text patterns." He stresses the importance: "These results show how advanced technologies such as BERT can help in early disease detection, not only for COVID-19, but also for other disease patterns and future outbreaks."
The dedicated research team continues to work hard to further develop and use these smart technologies to predict when infectious diseases may break out. This will allow GPs in the Netherlands to be better prepared to improve healthcare and be ready for future disease outbreaks. The promising results of this research were recently published in the Journal of Medical Internet Research (JMIR).