1. Computer vision methods under rapid evolution for pathology image tasks.
- Author
-
Maher, Nigel G, Scolyer, Richard A, and Liu, Sidong
- Subjects
- *
ARTIFICIAL intelligence , *CONVOLUTIONAL neural networks , *RGB color model , *SCHOLARSHIPS , *TRANSFORMER models , *DEEP learning - Abstract
The article discusses the rapid evolution of computer vision methods in the field of pathology, particularly focusing on artificial intelligence (AI) applications. It highlights the advancements in training deep-learning classifiers using convolutional neural networks (CNNs) and the emergence of self-supervised learning techniques to reduce the need for extensive annotations. The development of computational pathology foundation models, multiple-instance learning (MIL) frameworks, and innovations in annotation techniques are also explored. The article emphasizes the potential benefits of AI in pathology while acknowledging the importance of pathologist involvement and understanding of these evolving methods. [Extracted from the article]
- Published
- 2025
- Full Text
- View/download PDF