NSTL国家科技图书文献中心

1. Attention-Challenging Multiple Instance Learning for Whole Slide Image Classification NSTL国家科技图书文献中心

Yunlong Zhang | Honglin Li ... - 《Computer Vision - ECCV 2024,Part LIII》 - European Conference on Computer Vision - 2025, - 125～143 - 共19页

摘要：In the application of Multiple Instance | Learning (MIL) methods for Whole Slide Image (WSI | concentration. Firstly, UMAP of instance* features reveals* | . To remedy this, we introduce Multiple* Branch* | using multiple* attention branches. Secondly, the*

关键词： Computational pathology | Whole slide image | Multiple instance learning | Overfitting

2. cDP-MIL: Robust Multiple Instance Learning via Cascaded Dirichlet Process NSTL国家科技图书文献中心

Yihang Chen | Tsai Hor Chan ... - 《Computer Vision - ECCV 2024,Part LIV》 - European Conference on Computer Vision - 2025, - 232～250 - 共19页

摘要：Multiple instance learning (MIL) has been | framework for multiple* instance learning, which adopts a | approximate the true feature distribution of each instance | instance-to-bag characteristic of the WSIs. We perform | , which imposes a natural regularization on learning to*

关键词： Multiple instance learning | Whole slide images | Bayesian nonparametric method | Uncertainty estimation

3. DGR-MIL: Exploring Diverse Global Representation in Multiple Instance Learning for Whole Slide Image Classification NSTL国家科技图书文献中心

Wenhui Zhu | Xiwen Chen ... - 《Computer Vision - ECCV 2024,Part XXXVIII》 - European Conference on Computer Vision - 2025, - 333～351 - 共19页

摘要：Multiple instance learning (MIL) stands as a | learning paradigm. Specifically, the positive instance | powerful approach in weakly supervised learning | of all instances. First, we turn the instance | correlation into the similarity between instance* embeddings*

关键词： Weakly-supervised learning | Multiple instance learning | Histological whole slide image | Transformer

4. Multi-center Ovarian Tumor Classification Using Hierarchical Transformer-Based Multiple-Instance Learning NSTL国家科技图书文献中心

Cris H. B. Claessens | Eloy W. R. Schultz ... - 《Cancer Prevention, Detection, and Intervention》 - International Workshop on Cancer Prevention through Early Detection | International Conference on Medical Image Computing and Computer Assisted Intervention - 2025, - 3～13 - 共11页

摘要： multiple-instance learning (MIL) and hierarchical self | methods and other deep learning* approaches, the* | Malignant ovarian tumors (OTs) are a leading | cause of gynecological cancer deaths, and often remain | asymptomatic until advanced stages, making early and accurate

关键词： Ovarian cancer | Computer-aided diagnosis | Self-supervised learning | Multiple-instance learning | Vision transformer | External validation

5. Deep multiple instance learning on heterogeneous graph for drug–disease association prediction NSTL国家科技图书文献中心

Gu Y. | Zheng S.... - 《Computers in Biology and Medicine》 - 2025,184 - 109403～109403 - 共14页

摘要： this end, we introduce deep multiple instance | -end frameworks for path instance-level | representation learning* as well as the further feature* | learning into drug repositioning by proposing a novel | generator in MilGNet to obtain multiple* meta-path*

关键词： Drug repositioning | Drug–disease association prediction | Heterogeneous graph neural network | Meta-path | Multiple instance learning

6. Controlling false positives in multiple instance learning: The "c-rule" approach NSTL国家科技图书文献中心

Delgado, Rosario - 《International journal of approximate reasoning》 - 2025,179(Apr.) - 1.1～1.23 - 共23页

摘要： labeling bags in binary Multiple* Instance Learning (MIL* | approach addresses errors in instance* labeling by* | misclassifying a negative instance. This trend is further | This paper introduces a novel strategy for | ) under the standard MI assumption. The proposed

关键词： Multiple Instance Learning (MIL) | Standard MI assumption | False positive rate | False negative rate | LOGISTIC-REGRESSION

7. Refining Multiple Instance Learning with Attention Regularization for Whole Slide Image Classification NSTL国家科技图书文献中心

Ilan Carretero | Pablo Meseguer ... - 《Intelligent Data Engineering and Automated Learning - IDEAL 2024,Part I》 - International Conference on Intelligent Data Engineering and Automated Learning - 2025, - 398～409 - 共12页

摘要： multiple instance learning (MIL) problem where only slide | Histopathological analysis of biopsy sections | is crucial for the detection of cancer and the | distinction between different tumor subtypes. To this end | , pathologists identify certain key regions of the biopsy from

关键词： Hispathology imaging | Multiple instance learning | Attention mechanisms | Attention regularization | Explainable deep learning

8. Multiple-Instance Learning for thyroid gland disease classification: A hands-on experience NSTL国家科技图书文献中心

Lysukhin D. | Varlamov A.... - 《Computers in Biology and Medicine》 - 2025,184 - 109424～109424 - 共9页

摘要： datasets. Our study demonstrates that a Multiple-Instance | evaluating multiple* whole slide images (WSIs) from a single* | investigates the development of machine learning* models using* | Learning (MIL) model, trained on a weak patient-level | © 2024 Elsevier LtdThe morphological diagnosis

关键词： Artificial intelligence | Histopathology | Multiple Instance Learning | Thyroid cancer | Whole-slide images

9. Entity-level multiple instance learning for mesoscopic histopathology images classification with Bayesian collaborative learning and pathological prior transfer NSTL国家科技图书文献中心

He, Qiming | Xu, Yingming ... - 《Computerized Medical Imaging and Graphics》 - 2025,121 - 102495～102495 - 共12页

摘要： instances for multiple* instance learning. This restricts | a novel entity-level multiple instance* learning | proposed entity-level multiple* instance learning enables* | features, Bayesian collaborative learning* is proposed to* | construct co-optimization of instance* and bag embedding*

关键词： Pathology | Glomerular lesion pattern | Multiple instance learning | Mixup | Bayesian collaborative learning

10. Positional encoding-guided transformer-based multiple instance learning histopathology whole slide images classification NSTL国家科技图书文献中心

Shi, Jun | Sun, Dongdong ... - 《Computer Methods and Programs in Biomedicine》 - 2025,258 - 108491～108491 - 共14页

摘要： multiple instance learning (PEGTB-MIL) method for | attention. As the most representative, multiple instance | learning (MIL) generally aggregates the predictions or | Background and objectives: Whole slide image | (WSI) classification is of great clinical

摘要：In the application of Multiple Instance | Learning (MIL) methods for Whole Slide Image (WSI | concentration. Firstly, UMAP of instance* features reveals* | . To remedy this, we introduce Multiple* Branch* | using multiple* attention branches. Secondly, the*

关键词： Computational pathology | Whole slide image | Multiple instance learning | Overfitting

摘要：Multiple instance learning (MIL) has been | framework for multiple* instance learning, which adopts a | approximate the true feature distribution of each instance | instance-to-bag characteristic of the WSIs. We perform | , which imposes a natural regularization on learning to*

关键词： Multiple instance learning | Whole slide images | Bayesian nonparametric method | Uncertainty estimation

摘要：Multiple instance learning (MIL) stands as a | learning paradigm. Specifically, the positive instance | powerful approach in weakly supervised learning | of all instances. First, we turn the instance | correlation into the similarity between instance* embeddings*

关键词： Weakly-supervised learning | Multiple instance learning | Histological whole slide image | Transformer

摘要： multiple-instance learning (MIL) and hierarchical self | methods and other deep learning* approaches, the* | Malignant ovarian tumors (OTs) are a leading | cause of gynecological cancer deaths, and often remain | asymptomatic until advanced stages, making early and accurate

关键词： Ovarian cancer | Computer-aided diagnosis | Self-supervised learning | Multiple-instance learning | Vision transformer | External validation

摘要： this end, we introduce deep multiple instance | -end frameworks for path instance-level | representation learning* as well as the further feature* | learning into drug repositioning by proposing a novel | generator in MilGNet to obtain multiple* meta-path*

关键词： Drug repositioning | Drug–disease association prediction | Heterogeneous graph neural network | Meta-path | Multiple instance learning

摘要： labeling bags in binary Multiple* Instance Learning (MIL* | approach addresses errors in instance* labeling by* | misclassifying a negative instance. This trend is further | This paper introduces a novel strategy for | ) under the standard MI assumption. The proposed

关键词： Multiple Instance Learning (MIL) | Standard MI assumption | False positive rate | False negative rate | LOGISTIC-REGRESSION

摘要： multiple instance learning (MIL) problem where only slide | Histopathological analysis of biopsy sections | is crucial for the detection of cancer and the | distinction between different tumor subtypes. To this end | , pathologists identify certain key regions of the biopsy from

关键词： Hispathology imaging | Multiple instance learning | Attention mechanisms | Attention regularization | Explainable deep learning

关键词： Artificial intelligence | Histopathology | Multiple Instance Learning | Thyroid cancer | Whole-slide images

摘要： instances for multiple* instance learning. This restricts | a novel entity-level multiple instance* learning | proposed entity-level multiple* instance learning enables* | features, Bayesian collaborative learning* is proposed to* | construct co-optimization of instance* and bag embedding*

关键词： Pathology | Glomerular lesion pattern | Multiple instance learning | Mixup | Bayesian collaborative learning

摘要： multiple instance learning (PEGTB-MIL) method for | attention. As the most representative, multiple instance | learning (MIL) generally aggregates the predictions or | Background and objectives: Whole slide image | (WSI) classification is of great clinical

关键词： Digital pathology | Whole slide image | Multiple instance learning | Position encoding | Cancer subtyping | Gene mutation prediction

4008-161-200 800-990-8900

国家科技图书文献中心

摘要：In the application of Multiple Instance | Learning (MIL) methods for Whole Slide Image (WSI | concentration. Firstly, UMAP of instance features reveals | . To remedy this, we introduce Multiple Branch | using multiple attention branches. Secondly, the

关键词： Computational pathology | Whole slide image | Multiple instance learning | Overfitting

摘要：Multiple instance learning (MIL) has been | framework for multiple instance learning, which adopts a | approximate the true feature distribution of each instance | instance-to-bag characteristic of the WSIs. We perform | , which imposes a natural regularization on learning to

关键词： Multiple instance learning | Whole slide images | Bayesian nonparametric method | Uncertainty estimation

摘要：Multiple instance learning (MIL) stands as a | learning paradigm. Specifically, the positive instance | powerful approach in weakly supervised learning | of all instances. First, we turn the instance | correlation into the similarity between instance embeddings

关键词： Weakly-supervised learning | Multiple instance learning | Histological whole slide image | Transformer

摘要： multiple-instance learning (MIL) and hierarchical self | methods and other deep learning approaches, the | Malignant ovarian tumors (OTs) are a leading | cause of gynecological cancer deaths, and often remain | asymptomatic until advanced stages, making early and accurate

关键词： Ovarian cancer | Computer-aided diagnosis | Self-supervised learning | Multiple-instance learning | Vision transformer | External validation

摘要： this end, we introduce deep multiple instance | -end frameworks for path instance-level | representation learning as well as the further feature | learning into drug repositioning by proposing a novel | generator in MilGNet to obtain multiple meta-path

关键词： Drug repositioning | Drug–disease association prediction | Heterogeneous graph neural network | Meta-path | Multiple instance learning

摘要： labeling bags in binary Multiple Instance Learning (MIL | approach addresses errors in instance labeling by | misclassifying a negative instance. This trend is further | This paper introduces a novel strategy for | ) under the standard MI assumption. The proposed

关键词： Multiple Instance Learning (MIL) | Standard MI assumption | False positive rate | False negative rate | LOGISTIC-REGRESSION

摘要： multiple instance learning (MIL) problem where only slide | Histopathological analysis of biopsy sections | is crucial for the detection of cancer and the | distinction between different tumor subtypes. To this end | , pathologists identify certain key regions of the biopsy from

关键词： Hispathology imaging | Multiple instance learning | Attention mechanisms | Attention regularization | Explainable deep learning

摘要： datasets. Our study demonstrates that a Multiple-Instance | evaluating multiple whole slide images (WSIs) from a single | investigates the development of machine learning models using | Learning (MIL) model, trained on a weak patient-level | © 2024 Elsevier LtdThe morphological diagnosis

关键词： Artificial intelligence | Histopathology | Multiple Instance Learning | Thyroid cancer | Whole-slide images

摘要： instances for multiple instance learning. This restricts | a novel entity-level multiple instance learning | proposed entity-level multiple instance learning enables | features, Bayesian collaborative learning is proposed to | construct co-optimization of instance and bag embedding

关键词： Pathology | Glomerular lesion pattern | Multiple instance learning | Mixup | Bayesian collaborative learning

摘要： multiple instance learning (PEGTB-MIL) method for | attention. As the most representative, multiple instance | learning (MIL) generally aggregates the predictions or | Background and objectives: Whole slide image | (WSI) classification is of great clinical

关键词： Digital pathology | Whole slide image | Multiple instance learning | Position encoding | Cancer subtyping | Gene mutation prediction

4008-161-200 800-990-8900

国家科技图书文献中心

摘要：In the application of Multiple Instance | Learning (MIL) methods for Whole Slide Image (WSI | concentration. Firstly, UMAP of instance* features reveals* | . To remedy this, we introduce Multiple* Branch* | using multiple* attention branches. Secondly, the*

摘要：Multiple instance learning (MIL) has been | framework for multiple* instance learning, which adopts a | approximate the true feature distribution of each instance | instance-to-bag characteristic of the WSIs. We perform | , which imposes a natural regularization on learning to*

摘要：Multiple instance learning (MIL) stands as a | learning paradigm. Specifically, the positive instance | powerful approach in weakly supervised learning | of all instances. First, we turn the instance | correlation into the similarity between instance* embeddings*

摘要： multiple-instance learning (MIL) and hierarchical self | methods and other deep learning* approaches, the* | Malignant ovarian tumors (OTs) are a leading | cause of gynecological cancer deaths, and often remain | asymptomatic until advanced stages, making early and accurate

摘要： this end, we introduce deep multiple instance | -end frameworks for path instance-level | representation learning* as well as the further feature* | learning into drug repositioning by proposing a novel | generator in MilGNet to obtain multiple* meta-path*

摘要： labeling bags in binary Multiple* Instance Learning (MIL* | approach addresses errors in instance* labeling by* | misclassifying a negative instance. This trend is further | This paper introduces a novel strategy for | ) under the standard MI assumption. The proposed