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

Sinonasal Mucosal Epithelioid Melanoma with Rapid Skull-Base and Orbital Progression.

Journal of clinical medicine·2026
Same author

Classification of Microbial Activity and Inhibition Zones Using Neural Network Analysis of Laser Speckle Images.

Sensors (Basel, Switzerland)·2025
Same author

Oxidative Stress Modulation and Glutathione System Response During a 10-Day Multi-Stressor Field Training.

Journal of functional morphology and kinesiology·2025
Same author

Utilisation of Deep Neural Networks for Estimation of Cajal Cells in the Anal Canal Wall of Patients with Advanced Haemorrhoidal Disease Treated by LigaSure Surgery.

Cells·2025
Same author

AI-Assisted Detection and Localization of Spinal Metastatic Lesions.

Diagnostics (Basel, Switzerland)·2024
Same author

Foundational Models for Pathology and Endoscopy Images: Application for Gastric Inflammation.

Diagnostics (Basel, Switzerland)·2024

Related Experiment Video

Updated: May 26, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.6K

Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI.

Arturs Nikulins1, Edgars Edelmers1,2,3, Kaspars Sudars3

  • 1Faculty of Computer Science, Information Technology and Energy, Riga Technical University, LV-1048 Riga, Latvia.

Journal of Imaging
|February 25, 2025
PubMed
Summary

This study adapts classification neural networks for medical image segmentation, reducing the need for manual annotations. Explainable artificial intelligence (XAI) methods like GuidedBackprop effectively highlight anomalies in brain tumor datasets.

Keywords:
classification modelsexplainable artificial intelligenceimage segmentationmedical imagingneural networks

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

348
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.3K

Related Experiment Videos

Last Updated: May 26, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.6K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

348
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.3K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Segmentation neural networks are crucial for medical image analysis but require extensive annotated data and manual segmentation, posing privacy and efficiency challenges.
  • Classification neural networks capture essential features for object identification, offering a potential alternative for segmentation tasks.
  • Explainable Artificial Intelligence (XAI) techniques can provide insights into neural network decision-making processes.

Purpose of the Study:

  • To adapt classification neural networks for medical image segmentation, thereby reducing reliance on manual segmentation and annotated data.
  • To investigate the efficacy of XAI techniques in generating segmentation-like outputs from classification models.
  • To address data privacy concerns and improve the efficiency of medical image analysis.

Main Methods:

  • Utilized a ResNet classification neural network architecture.
  • Employed the Medical Segmentation Decathlon 'Brain Tumours' dataset for training and evaluation.
  • Applied various XAI tools, including GuidedBackprop, to generate segmentation-like heatmaps.

Main Results:

  • The adapted classification network successfully generated segmentation-like outputs.
  • GuidedBackprop demonstrated high efficiency and effectiveness, producing heatmaps that accurately highlighted target objects.
  • The approach reduced dependency on manual segmentation processes.

Conclusions:

  • Classification neural networks, when combined with XAI, offer a viable and efficient alternative for medical image segmentation.
  • XAI techniques, particularly GuidedBackprop, can bridge the gap between classification and segmentation tasks in medical imaging.
  • This method holds promise for overcoming data annotation limitations and privacy concerns in medical AI.