This presentation examines the development and application of Norwegian Clinical Large Language Models (LLMs) in healthcare. It explores how models like NorDeClin-BERT and NorBERT are reshaping key tasks such as clinical coding, adverse drug event detection, and the deidentification of clinical texts.
Significant challenges remain, including limited annotated data, privacy concerns, and algorithmic bias. At the same time, emerging trends like multimodal AI, federated learning, and synthetic data generation are shaping the future of the field.
With a focus on the continuous development of specialized Norwegian models and collaborative, ethically responsible approaches, this presentation highlights efforts to drive innovation, enhance healthcare delivery, and uphold patient safety and data privacy.
Read more and register through HIDA's webpages: