ASEF Young Minds
Online Events

Primož Kocbek: Predictive Models of Electronic Health Records

March 10, 2021 | 7:00 pm | Online

ASEF and the Student Club of Slovenske gorice are organizing an online popular science lecture on Predictive models of electronic health records, which will be conducted by Primož Kocbek, ASEF Fellow 2019 and doctoral student of biostatistics and research assistant at the University of Maribor, Faculty of Health Sciences. The lecture held in Slovene will take place online on Wednesday, March 10, 2021, at 7 PM CET.

To view the lecture, you must register by March 10, 2021, at the latest by 6 PM CET. The Zoom link for access and the code for asking questions will be sent to all registered users by e-mail on the day of the event. It is possible to register for the event here.

Electronic health records (EHI), defined as the longitudinal collection of health data about an individual patient or population, have become more interesting for use for secondary purposes or digitization analyzes outside the direct clinical treatment of the patient. Despite the very personal nature of the data, it is believed that such analyzes will contribute to the development of high-quality healthcare, better management and reduction of healthcare costs, and effective management of population health and clinical research. The lecture will present the basic characteristics and challenges in the construction of controlled forecasting models, such as the structure of the EHR data itself, as these are rare, high-volume, time-erratic data. Because it is very personal nature of the data, where some personal identifiers are masked or deidentified, it is necessary to strictly follow the protocols for access and use of them by the developer, as well as generally applicable legislation, such as, e.g., General regulation on data protection or GDPR in the EU. An example of a transparent study of the state of use of predictive EHR models in practice and the general need for high interpretability of predictive models in healthcare will be presented, showing some methods of global and local interpretability that can improve user confidence while fulfilling the GDPR right of interpretation (Right to Explanation).

Primož Kocbek is a graduate mathematician, doctoral student of biostatistics, and research assistant at the University of Maribor, Faculty of Health Sciences. In 2019, with the help of ASEF, he visited Stanford University, where he paid a research visit under the mentorship of computer science professor Jure Leskovac to predict the complications of chronic patients’ diseases based on EHR. Primož’s research interest includes statistical modeling and machine learning techniques with applications in healthcare.