Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Rethinking foundation models in pathology.

Nature biomedical engineering·2026
Same author

Artificial intelligence modelling in grading breast phyllodes tumours.

Histopathology·2026
Same author

Association of Kidney Volume With Patient-Reported Outcomes in ADPKD.

Kidney international reports·2026
Same author

Multimodal learning for scalable representation of high-dimensional medical data.

Frontiers in digital health·2026
Same author

Foreword.

Mayo Clinic proceedings. Digital health·2026
Same author

Magnetic Resonance Imaging for Distinguishing Perianal Hidradenitis Suppurativa from Fistulizing Crohn Disease.

Dermatology (Basel, Switzerland)·2025
Same journal

Krüppel-like factor 5 inhibition rescues cavernous nerve-injured erectile dysfunction by preventing phenotypic switch and mitochondrial dysfunction-dependent apoptosis in corpus cavernosum smooth muscle cells.

The American journal of pathology·2026
Same journal

APOE4 as a Modifier of Chemotherapy Response.

The American journal of pathology·2026
Same journal

Spatial Pathobiology in the Omics Era: Transforming Modern Pathology.

The American journal of pathology·2026
Same journal

Unbiased Spatial Proteomics Uncovers Hepatic in Situ Regulation in Alcohol-Associated Hepatitis.

The American journal of pathology·2026
Same journal

Zinc Finger Protein 143 Promotes Endometriotic Lesion Growth and Fibrosis through the Plasminogen Activator Inhibitor-1 Pathway.

The American journal of pathology·2026
Same journal

Dual Mechanisms Underlie the Repression of Repressor Element 1-Silencing Transcription Factor Expression in Lung Neuroendocrine Carcinoma Cells.

The American journal of pathology·2026
See all related articles

Related Experiment Video

Updated: Jul 4, 2025

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone
09:31

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone

Published on: April 8, 2015

11.6K

Model-Agnostic Binary Patch Grouping for Bone Marrow Whole Slide Image Representation.

Youqing Mu1, Hamid R Tizhoosh2, Taher Dehkharghanian3

  • 1Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, Ontario, Canada; Department of Pathology and Molecular Medicine, McMaster University, Hamilton, Ontario, Canada.

The American Journal of Pathology
|February 6, 2024
PubMed
Summary
This summary is machine-generated.

Binary patch grouping (BPG) enhances artificial intelligence models for analyzing whole slide images (WSIs) in bone marrow histopathology. This method improves WSI retrieval and classification accuracy by filtering less relevant image patches.

More Related Videos

DNA-barcode-based Multiplex Immunofluorescence Imaging to Analyze FFPE Specimens from Genetically Reprogrammed Murine Melanoma
09:52

DNA-barcode-based Multiplex Immunofluorescence Imaging to Analyze FFPE Specimens from Genetically Reprogrammed Murine Melanoma

Published on: June 6, 2025

195
Outer-Boundary Assisted Segmentation and Quantification of Trabecular Bones by an Imagej Plugin
09:36

Outer-Boundary Assisted Segmentation and Quantification of Trabecular Bones by an Imagej Plugin

Published on: March 14, 2018

9.4K

Related Experiment Videos

Last Updated: Jul 4, 2025

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone
09:31

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone

Published on: April 8, 2015

11.6K
DNA-barcode-based Multiplex Immunofluorescence Imaging to Analyze FFPE Specimens from Genetically Reprogrammed Murine Melanoma
09:52

DNA-barcode-based Multiplex Immunofluorescence Imaging to Analyze FFPE Specimens from Genetically Reprogrammed Murine Melanoma

Published on: June 6, 2025

195
Outer-Boundary Assisted Segmentation and Quantification of Trabecular Bones by an Imagej Plugin
09:36

Outer-Boundary Assisted Segmentation and Quantification of Trabecular Bones by an Imagej Plugin

Published on: March 14, 2018

9.4K

Area of Science:

  • Computational pathology
  • Digital histopathology
  • Artificial intelligence in medicine

Background:

  • Histopathology, the gold standard for disease diagnosis, faces challenges in efficiency and consistency with manual analysis of digitized whole slide images (WSIs).
  • Artificial intelligence (AI) models aim to create slide-level representations (single vectors) from WSIs to enable computational pathology tasks like search and classification.
  • Effective slide-level representations depend critically on patch feature extraction and aggregation strategies.

Purpose of the Study:

  • To introduce and evaluate a novel binary patch grouping (BPG) method as a plugin to enhance slide-level representations in bone marrow histopathology.
  • To investigate the impact of BPG on WSI retrieval and classification performance.
  • To compare domain-general versus domain-specific feature extraction models and different aggregation methods with and without BPG.

Main Methods:

  • Proposed a binary patch grouping (BPG) step, requiring minimal pathologist intervention, to exclude clinically irrelevant patches before feature aggregation.
  • Investigated both convolution- and attention-based feature extraction models, comparing domain-general and domain-specific approaches.
  • Evaluated two feature aggregation methods, with and without the BPG step, to assess its generalizability and impact on representation quality.

Main Results:

  • The BPG method demonstrated generalizability across different feature extraction and aggregation strategies.
  • Integration of BPG led to a 4% improvement in WSI retrieval performance (mean average precision at 10).
  • BPG resulted in a 5% improvement in WSI classification accuracy (weighted-F1 score) compared to pipelines without BPG.

Conclusions:

  • Binary patch grouping (BPG) is an effective plugin for enhancing the quality of slide-level representations in computational pathology.
  • The proposed BPG method improves the performance of key WSI analysis tasks, including retrieval and classification.
  • Domain-general large models combined with parameterized pooling offer the best performance for generating high-quality slide-level representations when using BPG.