Open Access

ARTICLE

Advanced Brain Tumor Segmentation in Magnetic Resonance Imaging via 3D U-Net and Generalized Gaussian Mixture Model-Based Preprocessing

Khalil Ibrahim Lairedj¹, Zouaoui Chama¹, Amina Bagdaoui¹, Samia Larguech², Younes Menni^3,4,*, Nidhal Becheikh⁵, Lioua Kolsi^6,*, Badr M. Alshammari⁷

1 Electronics, Photonics and Optronics Laboratory, Department of Electronics, Djillali Liabes University of Sidi Bel Abbes, Sidi Bel Abbes, 22000, Algeria
2 Department of Electrical Engineering, College of Engineering, Princess Nourah Bint Abdulrahman University, P.O. Box 84428, Riyadh, 11671, Saudi Arabia
3 Energy and Environment Laboratory, Department of Mechanical Engineering, Institute of Technology, University Center Salhi Ahmed Naama (Ctr. Univ. Naama), P.O. Box 66, Naama, 45000, Algeria
4 College of Technical Engineering, National University of Science and Technology, Dhi Qar, 64001, Iraq
5 Mining Research Center, Northern Border University, P.O. Box 1321, Arar, 91431, Saudi Arabia
6 Department of Mechanical Engineering, College of Engineering, University of Ha’il, Ha’il City, 81451, Saudi Arabia
7 Department of Electrical Engineering, College of Engineering, University of Ha’il, Ha’il City, 81451, Saudi Arabia

* Corresponding Authors: Younes Menni. Email: ; Lioua Kolsi. Email:

(This article belongs to the Special Issue: Advanced Computational Intelligence Techniques, Uncertain Knowledge Processing and Multi-Attribute Group Decision-Making Methods Applied in Modeling of Medical Diagnosis and Prognosis)

Computer Modeling in Engineering & Sciences 2025, 144(2), 2419-2443. https://doi.org/10.32604/cmes.2025.069396

Received 22 June 2025; Accepted 08 August 2025; Issue published 31 August 2025

Abstract

Brain tumor segmentation from Magnetic Resonance Imaging (MRI) supports neurologists and radiologists in analyzing tumors and developing personalized treatment plans, making it a crucial yet challenging task. Supervised models such as 3D U-Net perform well in this domain, but their accuracy significantly improves with appropriate preprocessing. This paper demonstrates the effectiveness of preprocessing in brain tumor segmentation by applying a pre-segmentation step based on the Generalized Gaussian Mixture Model (GGMM) to T1 contrast-enhanced MRI scans from the BraTS 2020 dataset. The Expectation-Maximization (EM) algorithm is employed to estimate parameters for four tissue classes, generating a new pre-segmented channel that enhances the training and performance of the 3D U-Net model. The proposed GGMM + 3D U-Net framework achieved a Dice coefficient of 0.88 for whole tumor segmentation, outperforming both the standard multiscale 3D U-Net (0.84) and MM U-Net (0.85). It also delivered higher Intersection over Union (IoU) scores compared to models trained without preprocessing or with simpler GMM-based segmentation. These results, supported by qualitative visualizations, suggest that GGMM-based preprocessing should be integrated into brain tumor segmentation pipelines to optimize performance.

---

Keywords

---