Saved in:
Bibliographic Details
Main Authors: Lai, Zhixin, Wu, Jing, Chen, Suiyao, Zhou, Yucheng, Hovakimyan, Naira
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.17343
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913289320005632
author Lai, Zhixin
Wu, Jing
Chen, Suiyao
Zhou, Yucheng
Hovakimyan, Naira
author_facet Lai, Zhixin
Wu, Jing
Chen, Suiyao
Zhou, Yucheng
Hovakimyan, Naira
contents In this study, we uncover the unexpected efficacy of residual-based large language models (LLMs) as part of encoders for biomedical imaging tasks, a domain traditionally devoid of language or textual data. The approach diverges from established methodologies by utilizing a frozen transformer block, extracted from pre-trained LLMs, as an innovative encoder layer for the direct processing of visual tokens. This strategy represents a significant departure from the standard multi-modal vision-language frameworks, which typically hinge on language-driven prompts and inputs. We found that these LLMs could boost performance across a spectrum of biomedical imaging applications, including both 2D and 3D visual classification tasks, serving as plug-and-play boosters. More interestingly, as a byproduct, we found that the proposed framework achieved superior performance, setting new state-of-the-art results on extensive, standardized datasets in MedMNIST-2D and 3D. Through this work, we aim to open new avenues for employing LLMs in biomedical imaging and enriching the understanding of their potential in this specialized domain.
format Preprint
id arxiv_https___arxiv_org_abs_2403_17343
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Residual-based Language Models are Free Boosters for Biomedical Imaging
Lai, Zhixin
Wu, Jing
Chen, Suiyao
Zhou, Yucheng
Hovakimyan, Naira
Computer Vision and Pattern Recognition
Computation and Language
Machine Learning
In this study, we uncover the unexpected efficacy of residual-based large language models (LLMs) as part of encoders for biomedical imaging tasks, a domain traditionally devoid of language or textual data. The approach diverges from established methodologies by utilizing a frozen transformer block, extracted from pre-trained LLMs, as an innovative encoder layer for the direct processing of visual tokens. This strategy represents a significant departure from the standard multi-modal vision-language frameworks, which typically hinge on language-driven prompts and inputs. We found that these LLMs could boost performance across a spectrum of biomedical imaging applications, including both 2D and 3D visual classification tasks, serving as plug-and-play boosters. More interestingly, as a byproduct, we found that the proposed framework achieved superior performance, setting new state-of-the-art results on extensive, standardized datasets in MedMNIST-2D and 3D. Through this work, we aim to open new avenues for employing LLMs in biomedical imaging and enriching the understanding of their potential in this specialized domain.
title Residual-based Language Models are Free Boosters for Biomedical Imaging
topic Computer Vision and Pattern Recognition
Computation and Language
Machine Learning
url https://arxiv.org/abs/2403.17343