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

Synthesis of 1,4,5,6-Tetrahydropyridazine Derivatives via Tandem Annulation of Coumarin Hydrazones and Bench-Stable Allenyl Sulfonium Salt.

Organic letters·2026
Same author

Deep Learning Technology in Genomics, Radiotherapy, and Ophthalmology for Precision Medicine.

Journal of physiological investigation·2026
Same author

Enantioselective Construction of Spirocyclohexenones via Ni-Catalyzed Double Arylative Cyclization of Alkyne-Tethered Malononitriles.

The Journal of organic chemistry·2025
Same author

An unconventional synthetic pathway for the reaction of alkynyl sulfonium ylides with aromatic acid anhydrides leading to the formation of sulfonium diacylmethylide frameworks.

Organic & biomolecular chemistry·2025
Same author

Reduction of respiratory motion artifacts in free-breathing abdominal MRI using strategic averaging of reassembled k-space data with Self-Modeled respiratory state sorting.

Physica medica : PM : an international journal devoted to the applications of physics to medicine and biology : official journal of the Italian Association of Biomedical Physics (AIFB)·2025
Same author

AI-Augmented Quantitative MRI Predicts Spontaneous Intracranial Hypotension.

Diagnostics (Basel, Switzerland)·2025

Related Experiment Video

Updated: Jun 15, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

379

Refining CT image analysis: Exploring adaptive fusion in U-nets for enhanced brain tissue segmentation.

Bang-Chuan Chen1, Chung-Yi Shen2, Jyh-Wen Chai3,4

  • 1The Department of Neurological Institute, Taichung Veterans General Hospital, Taichung, Taiwan, ROC.

Plos One
|June 11, 2025
PubMed
Summary

This study introduces an adaptive fusion strategy to improve deep-learning based infarction lesion segmentation, significantly reducing false alarms in brain imaging. The method enhances diagnostic accuracy for cerebral lesions.

More Related Videos

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

2.7K
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

2.4K

Related Experiment Videos

Last Updated: Jun 15, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

379
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

2.7K
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

2.4K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence in Medicine
  • Neurology

Background:

  • Non-contrast Computed Tomography (NCCT) is crucial for diagnosing acute cerebral hemorrhage or infarction.
  • Deep-learning (DL) algorithms for infarction lesion segmentation (ILS) can produce false alarms (FA) outside the brain region.

Purpose of the Study:

  • To develop an enhanced brain tissue segmentation method for ILS.
  • To reduce FAs in DL-based ILS by confining the search to cerebral tissue using an adaptive result fusion strategy.

Main Methods:

  • An adaptive result fusion strategy was integrated into DL-based ILS algorithms.
  • Various U-Net models were trained and evaluated with different fusion strategies.
  • A 9x9 Gaussian filter with unit standard deviation followed by binarization was used for refinement.
  • Performance was assessed using Intersection over Union (IoU) and Hausdorff Distance (HD) metrics, with external validation on the COCO dataset.

Main Results:

  • Fusion strategies using UNet2+ and UNet3+ achieved an IoU of 0.955 and HD of 1.33.
  • Fusion strategies involving U-net, UNet2+, and UNet3+ resulted in an IoU of 0.952 and HD of 1.61.
  • External validation on the COCO dataset showed IoU values around 0.46 and HD values around 584-728.

Conclusions:

  • The adaptive fusion strategy effectively reduces FAs and improves the training efficacy of DL-based ILS algorithms.
  • This data-independent methodology shows promise for versatile cerebral lesion segmentation.