Recently, several studies have reported on the fine-tuning of foundation models for image-text modeling in the field of medicine, utilizing images from online data sources such as Twitter and PubMed. Foundation models are large, deep artificial neural networks capable of learning the context of a specific domain through training on exceptionally extensive datasets. Through validation, we have observed that the representations generated by such models exhibit inferior performance in retrieval tasks within digital pathology when compared to those generated by significantly smaller, conventional deep networks.
@article{arxiv.2309.11510,
title = {When is a Foundation Model a Foundation Model},
author = {Saghir Alfasly and Peyman Nejat and Sobhan Hemati and Jibran Khan and Isaiah Lahr and Areej Alsaafin and Abubakr Shafique and Nneka Comfere and Dennis Murphree and Chady Meroueh and Saba Yasir and Aaron Mangold and Lisa Boardman and Vijay Shah and Joaquin J. Garcia and H. R. Tizhoosh},
journal= {arXiv preprint arXiv:2309.11510},
year = {2023}
}