Related Experiment Video
Updated: May 24, 2026
05:33
Three-Dimensional Shape Modeling and Analysis of Brain Structures
Published on: November 14, 2019
An Innovative 3D Slicer Plugin for Brain Images Annotation and Lesions Study
Marida de Maria1, Gianluigi Attanasio2, Martina de Salazar3
1Mediterranea University of Reggio Calabria, Italy.
Studies in Health Technology and Informatics
|May 23, 2026
Summary
A new plugin integrates deep learning models into 3D Slicer for biomedical image analysis. This tool enhances visualization and prediction accuracy using convolutional neural networks (CNNs).
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Computer-Aided Diagnosis
Background:
- Biomedical image analysis often requires specialized software.
- Integrating advanced machine learning models can improve diagnostic accuracy.
- Open-source platforms offer flexibility for research and development.
Purpose of the Study:
- To design and validate a plugin for 3D Slicer.
- To enable deep learning inference for biomedical image analysis within Slicer.
- To provide a seamless workflow for image processing, prediction, and visualization.
Main Methods:
- Developed a plugin integrating deep learning models into 3D Slicer.
- Implemented a VGG16-based convolutional neural network (CNN).
- Utilized ONNX Runtime for model inference within the Slicer framework.
- Employed pre-processing techniques including format conversion, normalization, and data augmentation.
Main Results:
- The plugin facilitates image loading, inference, and result visualization in 3D Slicer.
- Achieved approximately 85% accuracy and an Area Under the Curve (AUC) of 0.94.
- Demonstrated improved performance on high-definition medical images.
Conclusions:
- The developed plugin effectively integrates deep learning into 3D Slicer for biomedical imaging.
- The system offers a modular architecture for efficient image analysis workflows.
- This approach shows promise for enhancing computer-aided diagnosis in medical imaging.
