Overcoming data scarcity in biomedical imaging with a foundational multi-task model

IF 12 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS Nature computational science Pub Date : 2024-07-19 DOI:10.1038/s43588-024-00662-z
Raphael Schäfer, Till Nicke, Henning Höfener, Annkristin Lange, Dorit Merhof, Friedrich Feuerhake, Volkmar Schulz, Johannes Lotz, Fabian Kiessling
{"title":"Overcoming data scarcity in biomedical imaging with a foundational multi-task model","authors":"Raphael Schäfer, Till Nicke, Henning Höfener, Annkristin Lange, Dorit Merhof, Friedrich Feuerhake, Volkmar Schulz, Johannes Lotz, Fabian Kiessling","doi":"10.1038/s43588-024-00662-z","DOIUrl":null,"url":null,"abstract":"Foundational models, pretrained on a large scale, have demonstrated substantial success across non-medical domains. However, training these models typically requires large, comprehensive datasets, which contrasts with the smaller and more specialized datasets common in biomedical imaging. Here we propose a multi-task learning strategy that decouples the number of training tasks from memory requirements. We trained a universal biomedical pretrained model (UMedPT) on a multi-task database including tomographic, microscopic and X-ray images, with various labeling strategies such as classification, segmentation and object detection. The UMedPT foundational model outperformed ImageNet pretraining and previous state-of-the-art models. For classification tasks related to the pretraining database, it maintained its performance with only 1% of the original training data and without fine-tuning. For out-of-domain tasks it required only 50% of the original training data. In an external independent validation, imaging features extracted using UMedPT proved to set a new standard for cross-center transferability. UMedPT, a foundational model for biomedical imaging, has been trained on a variety of medical tasks with different types of label. It has achieved high performance with less training data in various clinical applications.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":null,"pages":null},"PeriodicalIF":12.0000,"publicationDate":"2024-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11288886/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Nature computational science","FirstCategoryId":"1085","ListUrlMain":"https://www.nature.com/articles/s43588-024-00662-z","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS","Score":null,"Total":0}
引用次数: 0

Abstract

Foundational models, pretrained on a large scale, have demonstrated substantial success across non-medical domains. However, training these models typically requires large, comprehensive datasets, which contrasts with the smaller and more specialized datasets common in biomedical imaging. Here we propose a multi-task learning strategy that decouples the number of training tasks from memory requirements. We trained a universal biomedical pretrained model (UMedPT) on a multi-task database including tomographic, microscopic and X-ray images, with various labeling strategies such as classification, segmentation and object detection. The UMedPT foundational model outperformed ImageNet pretraining and previous state-of-the-art models. For classification tasks related to the pretraining database, it maintained its performance with only 1% of the original training data and without fine-tuning. For out-of-domain tasks it required only 50% of the original training data. In an external independent validation, imaging features extracted using UMedPT proved to set a new standard for cross-center transferability. UMedPT, a foundational model for biomedical imaging, has been trained on a variety of medical tasks with different types of label. It has achieved high performance with less training data in various clinical applications.

Abstract Image

Abstract Image

查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
利用基础多任务模型克服生物医学成像中的数据匮乏问题。
经过大规模预训练的基础模型在非医疗领域取得了巨大成功。然而,训练这些模型通常需要大型、全面的数据集,这与生物医学成像中常见的更小、更专业的数据集形成了鲜明对比。在这里,我们提出了一种多任务学习策略,将训练任务的数量与内存要求分离开来。我们在一个多任务数据库上训练了一个通用生物医学预训练模型(UMedPT),该数据库包括断层扫描、显微镜和 X 射线图像,并采用了分类、分割和对象检测等多种标记策略。UMedPT 基础模型的表现优于 ImageNet 预训练模型和以前的先进模型。对于与预训练数据库相关的分类任务,只需使用 1%的原始训练数据,无需微调即可保持性能。对于域外任务,它只需要原始训练数据的 50%。在外部独立验证中,使用 UMedPT 提取的成像特征被证明是跨中心可转移性的新标准。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
CiteScore
11.70
自引率
0.00%
发文量
0
期刊最新文献
Real-time non-line-of-sight computational imaging using spectrum filtering and motion compensation. Deep generative design of RNA aptamers using structural predictions. Extracting reliable quantum outputs for noisy devices. Provable bounds for noise-free expectation values computed from noisy samples. E-waste challenges of generative artificial intelligence.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1