Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

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

PTCMIL: multiple instance learning via prompt token clustering for whole slide image analysis.

Medical image analysis·2026

Same author

Corrigendum to "LESS: Label-efficient multi-scale learning for cytological whole slide image screening" [Medical Image Analysis 94 (2024): 103109].

Medical image analysis·2025

Same author

AAV-delivered suppressor tRNA overcomes a nonsense mutation in mice.

Nature·2022

Same author

Surface plasmon enhancement in different spatial distributions of nanowires and two-dimensional materials.

Physical chemistry chemical physics : PCCP·2022

Same author

Zero-Field Splitting Calculations by Multiconfiguration Pair-Density Functional Theory.

Journal of chemical theory and computation·2022

Same author

Urinary metabolomics reveals the biological characteristics of early pregnancy in pigs.

Porcine health management·2022

Search research articles

Related Experiment Video

Updated: Jul 2, 2025

Automated Slide Scanning and Segmentation in Fluorescently-labeled Tissues Using a Widefield High-content Analysis System

Automated Slide Scanning and Segmentation in Fluorescently-labeled Tissues Using a Widefield High-content Analysis System

Published on: May 3, 2018

LESS: Label-efficient multi-scale learning for cytological whole slide image screening.

Beidi Zhao¹, Wenlong Deng¹, Zi Han Henry Li²

¹Department of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC V6T 1Z4, Canada; Vector Institute, Toronto, ON M5G 1M1, Canada.

Medical Image Analysis

|February 22, 2024

Summary

This summary is machine-generated.

We developed a Label-Efficient WSI Screening (LESS) method for analyzing cytological whole slide images (WSIs) using only slide-level labels. This approach improves accuracy and efficiency, particularly for small datasets, enabling automated cancer screening.

Keywords:

Computational pathology Multiple instance learning Whole slide image

More Related Videos

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Published on: October 4, 2024

Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research

Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research

Published on: June 21, 2024

Related Experiment Videos

Last Updated: Jul 2, 2025

Automated Slide Scanning and Segmentation in Fluorescently-labeled Tissues Using a Widefield High-content Analysis System

Automated Slide Scanning and Segmentation in Fluorescently-labeled Tissues Using a Widefield High-content Analysis System

Published on: May 3, 2018

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Published on: October 4, 2024

Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research

Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research

Published on: June 21, 2024

Area of Science:

Computational pathology
Digital pathology
Machine learning for medical imaging

Background:

Whole slide image (WSI) analysis in computational pathology faces computational challenges with large gigapixel images.
Existing multiple instance learning (MIL) methods for WSI analysis often overlook task-specific slide-level label supervision, leading to suboptimal feature extraction.
Pretrained or self-supervised models for patch feature extraction in WSIs can be ineffective or inefficient due to this oversight.

Purpose of the Study:

To propose a weakly-supervised, label-efficient method (LESS) for cytological WSI analysis, effective even with limited data.
To improve patch-level feature learning by incorporating slide-level labels for better task alignment.
To enhance WSI classification by addressing the sparse cell arrangement in cytological images.

Main Methods:

Employed variational positive-unlabeled (VPU) learning to infer hidden patch labels (benign/malignant) using slide-level supervision.
Implemented a multi-scale patch cropping strategy to capture information from sparse cellular arrangements.
Utilized a cross-attention vision transformer (CrossViT) to integrate multi-scale patch information for slide-level classification.

Main Results:

The LESS method achieved high performance on urine and breast cytology WSI datasets, with accuracy up to 96.88% and AUC up to 98.95%.
Demonstrated superior performance compared to state-of-the-art MIL methods on pathology WSIs.
Successfully enabled automated cytological WSI cancer screening.

Conclusions:

The proposed LESS method effectively addresses label-efficiency challenges in cytological WSI analysis.
Combining VPU learning with multi-scale CrossViT processing enhances feature extraction and classification accuracy.
LESS offers a promising solution for automated cancer screening using WSIs, especially in resource-limited scenarios.