The advent of foundation models marks a transformative era in artificial intelligence, with recent advancements unlocking unprecedented potential across diverse sectors such as healthcare, finance, and education. These large-scale pre-trained models, such as GPT-4 and BERT, have become the cornerstone of modern AI, offering versatile applications that range from natural language processing to computer vision. This talk explores how such models can be exploited for precision medicine. Firstly we will delve into traditional pathology methods and how AI can revolutionise the daily workflow of pathologists. By leveraging AI, pathologists can achieve more accurate and efficient diagnoses, thus improving patient care and outcomes. Secondly we will discuss the reliance on manual annotation for traditional AI pathology models, the associated challenges, and how foundation models may help to revolutionise this space.
Barbara Feulner is a Data Scientist @ Aignostics. In her role she led the development of a state-of-the-art foundation model for histopathology bringing major improvements to various downstream digital pathology tasks. She combines her background as a researcher studying neural networks with a passion for creating useful products in real-world applications. She is also curious about MLOps and the ethical aspects of modern AI.
