On November 27–29, prof. Vaidotas Marozas participated in the DAMSS 2025 Conference on Data Analysis Methods for Software Systems, organized by the Institute of Mathematics and Informatics of Vilnius University.
The event brought together researchers and practitioners working in data analysis, machine learning, signal processing, and software systems engineering.
During the conference, Vaidotas Marozas presented the study titled “Photoplethysmography-based Obstructive Sleep Apnea Detection Using ANN: Out-of-Distribution Study”, coauthors: Mantas Rinkevičius (Kaunas University of Technology), Oskar Pfeffer, Nando Hegemann (Physikalisch-Technische Bundesanstalt, Germany).
In this work, deep neural networks were investigated for the detection of obstructive sleep apnea from photoplethysmographic signals, with particular emphasis on out-of-distribution robustness — an essential requirement for the reliable deployment of machine-learning-based health monitoring solutions in both clinical environments and wearable-device applications.
The conference offered an excellent opportunity to discuss challenges and future directions in physiological signal analysis, domain shift problems, and AI applications in healthcare.
