Related Experiment Video
Updated: Apr 12, 2026
04:25
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
Published on: December 15, 2023
4.1K
Subcortical Region Segmentation using Fuzzy Based Augmented Lagrangian Multiphase Level Sets Method in Autistic MR
A R Jac Fredo1, G Kavitha, S Ramakrishnan
1Indian Institute of Technology Madras.
Biomedical Sciences Instrumentation
|May 22, 2015
Summary
This study uses an advanced imaging technique to segment brain regions in individuals with autism spectrum disorder (ASD) and controls. Findings reveal distinct texture patterns in brain regions of autistic subjects, potentially aiding in early diagnosis of neurodevelopmental disorders.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Computational Neuroscience
Background:
- Autism Spectrum Disorder (ASD) is a complex neurodevelopmental disorder.
- Accurate segmentation of brain regions is crucial for understanding structural differences.
- Existing segmentation methods may require complex initialization or re-initialization.
Purpose of the Study:
- To segment subcortical brain regions in control and autistic subjects using a novel imaging method.
- To analyze texture features (energy and entropy) in cortical and subcortical regions.
- To investigate the clinical significance of these features for ASD screening.
Main Methods:
- Utilized a Fuzzy C-Means (FCM) based Augmented Lagrangian (AL) multiphase level set method for brain segmentation.
- Employed FCM as an intensity discriminator within the multiphase level set framework.
- Validated segmentation accuracy using Dice Similarity Index and calculated texture features like energy and entropy.
Main Results:
- The multiphase level set method successfully segmented key subcortical regions (corpus callosum, brain stem, cerebellum).
- High Dice Similarity Index (above 0.85 for controls, 0.8 for autistic subjects) confirmed segmentation accuracy.
- Significant differences in energy and entropy were observed between control and autistic subjects, particularly in the brain stem and total brain regions.
Conclusions:
- The developed method offers robust segmentation of subcortical brain structures.
- Texture analysis revealed distinct patterns in autistic brains, with higher cortical energy and subcortical entropy.
- These findings suggest potential for early ASD detection and screening through neuroimaging biomarkers.
