Labeling instructions matter in biomedical image analysis
German Cancer Research Center (DKFZ)（德国癌症研究中心）；Heidelberg University（海德堡大学）；Helmholtz Zentrum München（亥姆霍兹慕尼黑中心）；Quality Match GmbH（Quality Match有限责任公司）
Tim Rädsch, Annika Reinke, Vivienn Weru, Minu D. Tizabi, Nicholas Schreck, A. Emre Kavur, Bünyamin Pekdemir, Tobias Roß, Annette Kopp-Schneider, Lena Maier-Hein
Biomedical image analysis algorithm validation depends on high-quality annotation of reference datasets, for which labeling instructions are key. Despite their importance, their optimization remains largely unexplored. Here, we present the first systematic study of labeling instructions and their impact on annotation quality in the field. Through comprehensive examination of professional practice and international competitions registered at the MICCAI Society, we uncovered a discrepancy between annotators' needs for labeling instructions and their current quality and availability. Based on an analysis of 14,040 images annotated by 156 annotators from four professional companies and 708 Amazon Mechanical Turk (MTurk) crowdworkers using instructions with different information density levels, we further found that including exemplary images significantly boosts annotation performance compared to text-only descriptions, while solely extending text descriptions does not. Finally, professional annotators constantly outperform MTurk crowdworkers. Our study raises awareness for the need of quality standards in biomedical image analysis labeling instructions.
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
Stanford University（斯坦福大学）；Emory University（埃默里大学）；University of California, Berkeley（加州大学伯克利分校）；University of Washington（华盛顿大学）；Massachusetts General Hospital（麻省总医院）；Harvard Medical School（哈佛医学院）
LLaDA-MedV: Exploring Large Language Diffusion Models for Biomedical Image Understanding
Arizona State University（亚利桑那州立大学）；Clemson University（克莱姆森大学）；LinkedIn Corporation（领英公司）；Washington University in St. Louis（圣路易斯华盛顿大学）；Morgan Stanley（摩根大通）
Learning by Correction: Efficient Tuning Task for Zero-Shot Generative Vision-Language Reasoning
School of Information Science and Technology, ShanghaiTech University（上海科技大学信息科学与技术学院）；Shanghai Engineering Research Center of Intelligent Vision and Imaging（上海智能视觉与成像工程技术研究中心）
From Compound Figures to Medical Multi-image Reasoning: Scaling Multimodal Large Language Models with Biomedical Literature
Department of Biomedical Informatics and Data Science, Yale School of Medicine, Yale University, New Haven, CT 06510, USA（耶鲁医学院生物医学信息学与数据科学系，耶鲁大学，新 Haven，CT 06510，USA）；School of Medicine, University of Puerto Rico, San Juan, PR 00921, USA（波多黎各大学医学院，波多黎各，San Juan，PR 00921，USA）
Finally Outshining the Random Baseline: A Simple and Effective Solution for Active Learning in 3D Biomedical Imaging
German Cancer Research Center (DKFZ)（德国癌症研究中心）；Helmholtz Imaging, German Cancer Research Center (DKFZ)（亥姆霍兹影像，德国癌症研究中心）；University of Heidelberg（海德堡大学）；ETH Zürich（苏黎世联邦理工学院）；Heidelberg University Hospital（海德堡大学医院）；National Center for Tumor Diseases (NCT) Heidelberg（海德堡国家肿瘤疾病中心）
nnActive: A Framework for Evaluation of Active Learning in 3D Biomedical Segmentation
German Cancer Research Center (DKFZ) Heidelberg, Division of Medical Image Computing（德国癌症研究中心（DKFZ）海德堡，医学影像计算部门）；Helmholtz Imaging, German Cancer Research Center (DKFZ), Heidelberg, Germany（海德堡影像学，德国癌症研究中心（DKFZ），海德堡，德国）；Faculty of Mathematics and Computer Science, University of Heidelberg, Germany（海德堡大学数学与计算机科学学院，德国）；German Cancer Research Center (DKFZ) Heidelberg, Division of Intelligent Medical Systems（德国癌症研究中心（DKFZ）海德堡，智能医学系统部门）；Institute for Machine Learning, ETH Zürich, Switzerland（苏黎世联邦理工学院机器学习研究所，瑞士）；German Cancer Research Center (DKFZ) Heidelberg, Interactive Machine Learning Group（德国癌症研究中心（DKFZ）海德堡，交互式机器学习小组）；Pattern Analysis and Learning Group, Department of Radiation Oncology, Heidelberg University Hospital, Germany（放射肿瘤学系模式分析与学习小组，海德堡大学医院，德国）；National Center for Tumor Diseases (NCT) Heidelberg, Germany（海德堡肿瘤疾病国家中心（NCT），德国）
Semi-supervised Medical Image Segmentation Method Based on Cross-pseudo Labeling Leveraging Strong and Weak Data Augmentation Strategies
Hangzhou Dianzi University（杭州电子科技大学）；Zhejiang University of Finance and Economics（浙江财经大学）；Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences（中国科学院深圳先进技术研究院）；Shenzhen Children's Hospital（深圳市儿童医院）；Shenzhen Research Institute of Big Data（深圳市大数据研究院）
VisualCheXbert: Addressing the Discrepancy Between Radiology Report Labels and Image Labels
Stanford University（斯坦福大学）；VinBrain
从自由文本放射学报告中自动提取医学状况对于监督计算机视觉模型解释医学图像至关重要。在这项工作中，我们表明，放射科医生标记报告显着不同意放射科医生标记相应的胸部X射线图像，这降低了质量的报告标签作为图像标签的代理。我们开发和评估的方法，以产生标签的放射学报告，有更好的协议与放射科医生标记图像。我们的最佳性能方法称为VisualCheXbert，它使用生物医学预训练的BERT模型直接从放射学报告映射到图像标签，并通过训练用于从胸部X射线图像中检测医疗状况的计算机视觉模型确定监督信号。我们发现，VisualCheXbert优于使用现有放射学报告标签的方法，平均F1得分为0.14（95%CI 0.12，0.17）。我们还发现，VisualCheXpert更好地同意放射科医生标记胸部X射线图像比放射科医生标记相应的放射学报告之间的平均F1分数在0.12（95% CI 0.09，0.15）和0.21（95% CI 0.18，0.24）之间的几种医疗条件。
Pseudo-Labeling for Small Lesion Detection on Diabetic Retinopathy Images
Department of Computer Science University of Massachusetts Lowell（马萨诸塞大学洛厄尔分校计算机科学系）；Department of Ophthalmology Central South University（中南大学湘雅医学院（注：原文仅显示Department of Ophthalmology，此处“Central South University”为大学，按要求提取到大学级别））
