2026-05-09T16:08:43.408Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:1:10095 2024-12-03T20:10:47Z 1

Generative Adversarial Network for Data Augmentation and Substitution 2024 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/1/10095 https://ieeexplore.ieee.org/abstract/document/10579362 M. Stankovic N. Bacanin M. Zivkovic L. Jovanovic M. Sarac M. Antonijevic The healthcare industry undergoes a transformative shift with the integration of artificial intelligence (AI), particularly in medical imaging analysis. AI algorithms, trained on extensive datasets, demonstrate accuracy comparable to medical experts, promising earlier diagnoses and improved outcomes. However, challenges in widespread AI adoption include limited access to high-quality medical datasets and concerns about data privacy. These challenges hinder progress, as existing datasets often suffer from obsolescence. Overcoming these limitations is crucial for advancing AI research in healthcare. Leveraging Generative Adversarial Networks (GANs) addresses data scarcity by generating synthetic medical data that maintains essential features without compromising patient privacy. GANs present a solution to challenges related to data availability and quality, unlocking new opportunities for innovation in medical AI research and clinical practice. conferenceObject 7 12 10.1109/ZINC61849.2024.10579362 2024 Zooming Innovation in Consumer Technologies Conference (ZINC), IEEE, Novi Sad, Serbia