TriDeNT : Triple deep network training for privileged knowledge distillation in histopathology

IF 10.7 1区医学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Medical image analysis Pub Date : 2025-03-18 DOI:10.1016/j.media.2025.103479

Lucas Farndale , Robert Insall , Ke Yuan

{"title":"TriDeNT : Triple deep network training for privileged knowledge distillation in histopathology","authors":"Lucas Farndale , Robert Insall , Ke Yuan","doi":"10.1016/j.media.2025.103479","DOIUrl":null,"url":null,"abstract":"<div><div>Computational pathology models rarely utilise data that will not be available for inference. This means most models cannot learn from highly informative data such as additional immunohistochemical (IHC) stains and spatial transcriptomics. We present TriDeNT <figure><img></figure>, a novel self-supervised method for utilising privileged data that is not available during inference to improve performance. We demonstrate the efficacy of this method for a range of different paired data including immunohistochemistry, spatial transcriptomics and expert nuclei annotations. In all settings, TriDeNT <figure><img></figure> outperforms other state-of-the-art methods in downstream tasks, with observed improvements of up to 101%. Furthermore, we provide qualitative and quantitative measurements of the features learned by these models and how they differ from baselines. TriDeNT <figure><img></figure> offers a novel method to distil knowledge from scarce or costly data during training, to create significantly better models for routine inputs.</div></div>","PeriodicalId":18328,"journal":{"name":"Medical image analysis","volume":"102 ","pages":"Article 103479"},"PeriodicalIF":10.7000,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Medical image analysis","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1361841525000271","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}

引用次数: 0

Abstract

Computational pathology models rarely utilise data that will not be available for inference. This means most models cannot learn from highly informative data such as additional immunohistochemical (IHC) stains and spatial transcriptomics. We present TriDeNT

, a novel self-supervised method for utilising privileged data that is not available during inference to improve performance. We demonstrate the efficacy of this method for a range of different paired data including immunohistochemistry, spatial transcriptomics and expert nuclei annotations. In all settings, TriDeNT

outperforms other state-of-the-art methods in downstream tasks, with observed improvements of up to 101%. Furthermore, we provide qualitative and quantitative measurements of the features learned by these models and how they differ from baselines. TriDeNT

offers a novel method to distil knowledge from scarce or costly data during training, to create significantly better models for routine inputs.

查看原文本刊更多论文

三叉戟：组织病理学专有知识蒸馏的三重深度网络训练

计算病理学模型很少使用不能用于推理的数据。这意味着大多数模型不能从高信息量的数据中学习，如额外的免疫组化（IHC）染色和空间转录组学。我们提出了TriDeNT，一种新的自监督方法，用于利用在推理过程中不可用的特权数据来提高性能。我们证明了这种方法对一系列不同配对数据的有效性，包括免疫组织化学、空间转录组学和专家核注释。在所有情况下，TriDeNT在下游任务中的表现都优于其他最先进的方法，改进幅度高达101%。此外，我们提供了这些模型所学习的特征的定性和定量测量，以及它们与基线的区别。TriDeNT提供了一种新颖的方法，可以在训练期间从稀缺或昂贵的数据中提取知识，从而为日常输入创建更好的模型。

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分钟内获得全文求助全文

来源期刊

Medical image analysis 工程技术-工程：生物医学

CiteScore

22.10

自引率

6.40%

发文量

309

审稿时长

6.6 months

期刊介绍： Medical Image Analysis serves as a platform for sharing new research findings in the realm of medical and biological image analysis, with a focus on applications of computer vision, virtual reality, and robotics to biomedical imaging challenges. The journal prioritizes the publication of high-quality, original papers contributing to the fundamental science of processing, analyzing, and utilizing medical and biological images. It welcomes approaches utilizing biomedical image datasets across all spatial scales, from molecular/cellular imaging to tissue/organ imaging.