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

Detailed Structure and Function of Lymph Nodes01:23

Detailed Structure and Function of Lymph Nodes

2.9K
Lymph nodes are bean-shaped structures that cluster along the lymphatic vessels in the inguinal, axillary, and cervical regions. Each node is divided into compartments by a capsule that extends trabeculae inward.
From a histological perspective, lymph nodes can be split into two main areas: the superficial cortex and the deep medulla. The outer cortex is populated by dendritic cells, macrophages, and B lymphocytes, which are densely packed into follicles. When these B-lymphocytes are presented...
2.9K
Metastasis02:30

Metastasis

5.8K
Metastasis is the spread of cancer cells from the original site to distant locations in the body. Cancer cells can spread via blood vessels (hematogenous) as well as lymph vessels in the body.
Epithelial-to-Mesenchymal Transition
The epithelial-to-mesenchymal transition or EMT is a developmental process commonly observed in wound healing, embryogenesis, and cancer metastasis. EMT is induced by transforming growth factor-beta (TGF-β) or receptor tyrosine kinase (RTK) ligands, which further...
5.8K
Classification of Leukocytes01:30

Classification of Leukocytes

3.6K
Leukocytes are classified into two groups based on the presence or absence of cytoplasmic granules. Granular leukocytes, which contain granules, belong to the myeloid lineage and are divided into three subtypes: neutrophils, eosinophils, and basophils. These cells are roughly spherical and characterized by the granules in their cytoplasm.
Neutrophils are the most abundant type of granular leukocytes, comprising 50-70% of all leukocytes. They feature small, evenly distributed granules and a...
3.6K

You might also read

Related Articles

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

Sort by
Same author

Explainability Methods for AI-Assisted Diagnosis of Lymph Node Metastases in Digital Pathology: A Quantitative Comparative Study.

Diagnostics (Basel, Switzerland)·2026
Same author

Explainable Artificial Intelligence for Bias Detection in COVID CT-Scan Classifiers.

Sensors (Basel, Switzerland)·2021
Same author

Convolutional ensembles for Arabic Handwritten Character and Digit Recognition.

PeerJ. Computer science·2021
Same author

Local Interpretable Model-Agnostic Explanations for Classification of Lymph Node Metastases.

Sensors (Basel, Switzerland)·2019
Same author

Safe exposure distances for transcranial magnetic stimulation based on computer simulations.

PeerJ·2018
Same author

Evolving Spiking Neural Networks for Recognition of Aged Voices.

Journal of voice : official journal of the Voice Foundation·2016

Related Experiment Video

Updated: Oct 6, 2025

Quantification of Tumor Cell Adhesion in Lymph Node Cryosections
06:09

Quantification of Tumor Cell Adhesion in Lymph Node Cryosections

Published on: February 9, 2020

12.5K

Evolved explainable classifications for lymph node metastases.

Iam Palatnik de Sousa1, Marley M B R Vellasco1, Eduardo Costa da Silva1

  • 1Pontifical Catholic University of Rio de Janeiro, Marquês de São Vicente Street, 22541-041, 225 - Gávea, Rio de Janeiro, Brazil.

Neural Networks : the Official Journal of the International Neural Network Society
|January 19, 2022
PubMed
Summary

This study introduces Evolved Explanations (EvEx), an AI method combining LIME and genetic algorithms for automated tuning in image classification. EvEx generates reliable explanations for medical image analysis, aligning with expert segmentations.

Keywords:
Artificial intelligenceConvolutional Neural NetworksExplainable AIMulti-objective genetic algorithms

More Related Videos

Author Spotlight: A Model to Study the Systemic and Local Dynamics of CD8+ T Cells During LN Metastasis
07:45

Author Spotlight: A Model to Study the Systemic and Local Dynamics of CD8+ T Cells During LN Metastasis

Published on: January 26, 2024

2.2K
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
07:13

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model

Published on: April 18, 2025

239

Related Experiment Videos

Last Updated: Oct 6, 2025

Quantification of Tumor Cell Adhesion in Lymph Node Cryosections
06:09

Quantification of Tumor Cell Adhesion in Lymph Node Cryosections

Published on: February 9, 2020

12.5K
Author Spotlight: A Model to Study the Systemic and Local Dynamics of CD8+ T Cells During LN Metastasis
07:45

Author Spotlight: A Model to Study the Systemic and Local Dynamics of CD8+ T Cells During LN Metastasis

Published on: January 26, 2024

2.2K
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
07:13

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model

Published on: April 18, 2025

239

Area of Science:

  • Artificial Intelligence
  • Medical Image Analysis
  • Explainable AI

Background:

  • Deep learning models, particularly Convolutional Neural Networks (CNNs), are powerful for medical image classification but often function as 'black boxes'.
  • Interpreting the decisions of these models is crucial for clinical adoption and trust, especially in tasks like identifying metastatic tissue in pathology slides.

Purpose of the Study:

  • To present a novel evolutionary approach, Evolved Explanations (EvEx), for enhancing Explainable Artificial Intelligence (XAI) in medical image classification.
  • To automate the tuning of segmentation parameters within the Local Interpretable Model-Agnostic Explanations (LIME) framework using Multi-Objective Genetic Algorithms.

Main Methods:

  • The EvEx model integrates LIME with Multi-Objective Genetic Algorithms for automated parameter optimization.
  • A CNN was trained on the Patch-Camelyon dataset for binary classification of lymph node metastatic tissue.
  • Explanations were generated by evolving segmentations to simultaneously optimize three evaluation goals, with the final explanation being the mean of Pareto front solutions.

Main Results:

  • The developed genetic algorithm successfully optimized segmentation parameters, leading to high-quality explanations.
  • Explanations generated from different random seeds showed remarkable agreement, indicating reproducibility.
  • The resulting heat maps from EvEx aligned well with expert medical segmentations, validating the method's accuracy.

Conclusions:

  • The EvEx methodology provides a robust and automated approach to generating reliable explanations for CNN-based medical image classification.
  • This technique offers valuable insights into neural network decision-making processes for pathology images.
  • The automated parameter tuning represents a significant advancement for XAI in clinical applications.