Related Experiment Video
Updated: Jul 16, 2026
14:15
Multicolor 3D Printing of Complex Intracranial Tumors in Neurosurgery
Published on: January 11, 2020
Modeling of intensity priors for knowledge-based level set algorithm in calvarial tumors segmentation.
Aleksandra Popovic1, Ting Wu, Martin Engelhardt
1Chair of Medical Engineering, Helmholtz-Institute for Biomedical Engineering, RWTH Aachen University, Germany. popovic@hia.rwth-aachen.de
Medical Image Computing and Computer-Assisted Intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
|March 16, 2007
Summary
This study introduces an automated framework for segmenting 3D calvarial tumors using Computed Tomography scans. The research validates various intensity modeling approaches for improved tumor segmentation accuracy.
Area of Science:
- Medical Imaging
- Computational Biology
- Image Analysis
Background:
- Automatic segmentation of 3D calvarial tumors from Computed Tomography (CT) images is challenging due to tumors spanning soft and bone tissues with diverse intensity ranges.
- Accurate segmentation is crucial for computational modeling and treatment planning of calvarial tumors.
Purpose of the Study:
- To present an automatic knowledge-based framework for level set segmentation of 3D calvarial tumors.
- To analyze and validate different intensity prior modeling approaches for multiclass segmentation problems.
- To evaluate the accuracy of Gaussian mixture models (one, two, and three class) and a discrete model.
Main Methods:
- Development of an automatic knowledge-based framework utilizing level set segmentation.
- Evaluation of one, two, and three class Gaussian mixture models and a discrete model for intensity prior modeling.
- Validation of segmentation results against manually segmented golden standards using Receiver Operating Curve (ROC) analysis and Dice similarity coefficient.
Main Results:
- Comparison of probability density modeling accuracy across different models.
- Assessment of segmentation outcome based on validated metrics.
- Demonstration of the framework's capability in segmenting complex calvarial tumors.
Conclusions:
- The presented framework offers a robust solution for automatic 3D calvarial tumor segmentation.
- The study provides insights into the effectiveness of various intensity prior models for multiclass segmentation tasks.
- Validated segmentation results support the clinical utility of the proposed approach for computational modeling.
