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 Experiment Videos

Fully automated biomedical image segmentation by self-organized model adaptation.

Axel Wismüller1, Frank Vietze, Johannes Behrends

  • 1Institut für Klinische Radiologie, Ludwig-Maximilians-Universität München, Klinikum Innenstadt, Ziemessenstrasse 1, München 80336, Germany. axel@wismueller.de

Neural Networks : the Official Journal of the International Neural Network Society
|November 24, 2004
PubMed
Summary

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

[Sarcomas-Red flags for radiology].

Radiologie (Heidelberg, Germany)·2026
Same author

Multimodal contrastive prognostication framework for early neurological outcome prediction in post-cardiac arrest patients.

Scientific reports·2026
Same author

Inferring causal relations from multivariate data using Large-Scale Augmented Granger Causality (lsAGC).

NeuroImage·2025
Same author

Impact of sensor configuration and melanin concentration on reflective pulse oximetry using Monte Carlo simulations.

Scientific reports·2025
Same author

[Ghosts and half-shadows-Shedding light on the mysterious world of inflammations and infections of the bone].

Radiologie (Heidelberg, Germany)·2025
Same author

Electric-Field Molecular Fingerprinting to Probe Cancer.

ACS central science·2025

This study introduces a fully automated image segmentation method using the deformable (feature) map (DM) algorithm. This approach enables efficient, adaptive learning for multispectral brain MRI segmentation without manual intervention.

Area of Science:

  • Biomedical Image Processing
  • Machine Learning
  • Computational Neuroscience

Background:

  • Accurate segmentation of multispectral magnetic resonance imaging (MRI) data is crucial for analyzing human brain structures.
  • Traditional segmentation methods often require manual intervention, limiting efficiency and reproducibility.
  • Developing automated, adaptive segmentation techniques is essential for advancing biomedical image analysis.

Purpose of the Study:

  • To present a fully automated image segmentation method using the deformable (feature) map (DM) algorithm.
  • To demonstrate the algorithm's capability for adaptive learning and re-utilization of knowledge from reference datasets.
  • To apply the method to multispectral human brain MRI data for voxel-based segmentation.

Main Methods:

Related Experiment Videos

  • Utilized the deformable (feature) map (DM) algorithm for function approximation and adaptive plasticity.
  • Reduced segmentation problems to supervised one-shot training on a single dataset.
  • Implemented a similarity transformation based on self-organized deformation of probability distributions.
  • Applied incremental adaptive learning for segmenting new, similar datasets.

    • Main Results:

    • Achieved fully automated voxel-based multispectral image segmentation of human brain MRI data.
    • Demonstrated that knowledge from a reference dataset can be re-utilized for segmenting new data.
    • Eliminated the need for manual contour tracing, visual classification, or human intervention.
    • Showcased the efficiency and practicability of the automated segmentation system.

    Conclusions:

    • The deformable (feature) map (DM) algorithm provides a robust solution for fully automated image segmentation.
    • Self-organized incremental model adaptation enhances the efficiency and practicality of biomedical image processing.
    • This automated approach has significant implications for reproducible and scalable neuroimaging research.