Main Functions in GlioMap

1Load Module: Load the MRI Image

GlioMap supports various file formats for comprehensive medical image analysis:

Supported Formats: NIfTI (.nii, .nii.gz), MetaImage (.mhd, .mha), NRRD (.nrrd, .nhdr), DICOM (.dcm, .dicom), Analyze (.hdr, .img), FreeSurfer (.mgz, .mgh), GIPL (.gipl, .gipl.gz), VTK (.vtk, .vti, .vtp, .vtu), MINC (.mnc, .minc), PIC, LSM, and standard image formats (.tif, .png, .jpg, etc.).

2Skull Strip Module: Remove Skull Area from Images

GlioMap uses a simplified version of the BET (Brain Extraction Tool) algorithm for skull stripping. The process involves several mathematical operations:

Brain Extraction Algorithm Steps
Normalization:
\( I_{\mathrm{norm}}(x,y) = \frac{I(x,y) - \min(I)}{\max(I) - \min(I)} \)
Gaussian Smoothing:
\( G(x,y) = \frac{1}{2\pi\sigma^2} \exp\left(-\frac{x^2 + y^2}{2\sigma^2}\right) \)
Smoothed Image Convolution:
\( I_{\mathrm{smooth}}(x,y) = I_{\mathrm{norm}}(x,y) \ast G(x,y) \)
Intensity Thresholding:
\( T = \mathrm{threshold} \cdot \max(I_{\mathrm{smooth}}) \)
Morphological Operations:
Opening: \( M_{\mathrm{open}} = (M_{\mathrm{initial}} \ominus B) \oplus B \)
Closing: \( M_{\mathrm{close}} = (M_{\mathrm{open}} \oplus B) \ominus B \)
Largest Connected Component Selection:
\( C_{\mathrm{max}} = \arg\max_i \left| C_i \right| \)
Final Smoothing:
\( M_{\mathrm{final}} = G_{\sigma=1} \ast M_{\mathrm{close}} \)
Binarization:
\( M_{\mathrm{binary}}(x,y) = \left[ \left( G_{\sigma=1} \ast M_{\mathrm{close}} \right)(x,y) > 0.5 \right] \)

3Preprocessing Module: Normalization and Bias Field Correction

Provides essential preprocessing capabilities including intensity normalization and bias field correction to enhance image quality and standardize pixel intensities across different acquisitions.

4Slice Navigator: Navigate Through Slice Sequences

Facilitates navigating through the slice sequence and playing the sequence as a video for comprehensive 3D visualization and analysis.

5Contrast and Brightness Changer: Visual Enhancement

Facilitates changing contrast and brightness parameters for better visualization and improved identification of anatomical structures and pathological regions.

6Select Slice Module: Tumor Slice Selection

Facilitates selecting specific tumor slices in the sequence for focused analysis and feature extraction.

7Thresholding Tools: MRI Image Thresholding

Advanced thresholding capabilities using Fuzzy C-Means (FCM) clustering for precise tumor segmentation:

Fuzzy C-Means Thresholding Algorithm
Input Image Representation:
\( X = \{x_1, x_2, \ldots, x_N\},\quad \mathrm{where\ } N = R \cdot C \)
Fuzzy C-Means Objective Function:
\( J = \sum_{i=1}^{N} \sum_{j=1}^{C} u_{ji}^m \, \lVert x_i - v_j \rVert^2 \)
Membership Update:
\( u_{ji} = \frac{1}{\sum_{k=1}^{C} \left( \frac{\lVert x_i - v_j \rVert}{\lVert x_i - v_k \rVert} \right)^{\frac{2}{m-1}}} \)
Cluster Center Update:
\( v_j = \frac{\sum_{i=1}^{N} u_{ji}^m \, x_i}{\sum_{i=1}^{N} u_{ji}^m} \)
Tumor Membership Map:
\( U_{\mathrm{tumor}}(x,y) = \mathrm{reshape}(u_{c,i}, R, C),\quad \mathrm{where\ } c = \arg\max(v_j) \)
Adaptive Thresholding:
\( T = \mathrm{adaptiveThreshold}(U_{\mathrm{tumor}}) \)
Post-processing (Morphology):
Closing: \( M_{\mathrm{closed}} = (M \oplus B) \ominus B \)
Area Filtering: \( M_{\mathrm{filtered}} = \mathrm{remove\ small\ components\ < 50\ pixels} \)
Hole Filling: \( M_{\mathrm{final}} = \mathrm{fill\ holes\ in\ } M_{\mathrm{filtered}} \)
Output:
\( M_{\mathrm{final}}(x,y) = \text{Segmented tumor region} \)

8Draw and Erase Tools: Manual Segmentation

Interactive tools that facilitate drawing or erasing threshold areas for manual refinement of automated segmentation results.

9ROI Manager: Region of Interest Management

Comprehensive tools for drawing, adjusting, and erasing ROI areas for precise tumor segmentation and region-specific analysis.

10Masking Module: Mask Generation and Application

Facilitates generating binary mask images and applying masks to other images for focused analysis of specific anatomical regions.

11Feature Calculation Module: Radiomic and Morphological Analysis

Advanced feature extraction capabilities for calculating comprehensive radiomic and morphological features for the entire MRI image or selected regions of interest.

12Output Module: Data Export and Saving

Comprehensive save options for segmented images, skull-stripped images, mask images, and calculated features in various standard formats for further analysis and research.