Brain tumours are seen as critical health challenges that require accurate and early detection for effective treatment. An evaluation of two unsupervised machine learning models Gaussian mixture Fuzzy C Means ensemble model and Spectral Clustering to see the potential use case for potential medical applications. Both models evaluated 800 samples dataset of MRI images using 80 percent of data as training with 20 percent left as testing. The Gaussian Mixture - Fuzzy C Means showed a perfect prediction of all gliomas as opposed to Spectral encountering a catastrophic failure in the “No Tumour” class where all were misclassified as other types of tumours.
Table of Contents
1. Abstract
2. Introduction
3. Literature Review
4. Methodology
4.1 MRI Brain Dataset and Preprocessing
4.2 Feature Extraction
4.2.1 Intensity features
4.2.2 Gradient Texture
4.2.3 Morphological Features
4.2.4 Frequency Features
4.2.5 Statistics
4.2.6 Edge Detection
4.2.7 Histogram
4.3 Gaussian Mixture Model and Fuzzy C Means Ensemble Model (GMM-FCM)
4.4 Spectral Clustering Model including Multiscale Analysis
4.5 Alignment Of Clusters and Evaluation
5. Results and Discussion
5.1 Model Performance
5.2 Models Confusion Matrix
5.3 ROC-AUC Comparison
6. Conclusion
7. References
Research Objectives and Key Topics
The primary objective of this research is to evaluate and compare the effectiveness of two unsupervised machine learning models, the Gaussian Mixture-Fuzzy C Means (GMM-FCM) ensemble and Spectral Clustering, in the classification of brain tumours from MRI imagery. The study aims to determine which model offers superior diagnostic potential for clinical applications, particularly when working with unlabelled datasets.
- Comparative analysis of unsupervised machine learning models.
- Application of MRI image processing and feature extraction techniques.
- Evaluation of GMM-FCM ensemble performance vs. Spectral Clustering.
- Assessment of diagnostic accuracy in medical imaging.
- Identification of limitations in unsupervised brain tumour detection.
Excerpt from the Book
Gaussian Mixture Model and Fuzzy C Means Ensemble Model (GMM-FCM)
The GMM-FCM ensemble model is a hybrid approach of probabilistic and geometric clustering. The Gaussian Mixture component will assume the datapoints that are created from distributions of multiple Gaussian models with each one representing a tumor class.
The Fuzzy C Means Component introduces soft clustering as each image is given a degree of membership to all clusters as opposed to belonging to only one cluster. This is valuable in imaging where brain tumours can be unclear with mixed characteristics.
The model functions by updating assigned clusters through two processes. The expectation will compute the membership probabilities using GMM’s likelihood and FCM’s distance assignments. Maximization will update the centres of each cluster and mixing weights based on these connections. Hybrid parameter (Gamma =6) controls how much will be the model balance between GMM and FCM with this parameter favouring GMM and using FCM to outliers.
Summary of Chapters
Abstract: Summarizes the study’s focus on comparing two unsupervised machine learning models for brain tumour detection using MRI data and highlights the superior performance of the GMM-FCM ensemble.
Introduction: Provides context on the clinical importance of early brain tumour diagnosis and outlines the research intention to test unsupervised models on unlabelled datasets.
Literature Review: Discusses existing medical imaging techniques and previous applications of GMM, Spectral Clustering, and Fuzzy C Means in healthcare and anomaly detection.
Methodology: Details the dataset preparation, the seven-category feature extraction process, and the specific architecture of the GMM-FCM and Spectral Clustering models.
Results and Discussion: Presents the comparative performance metrics and confusion matrices, demonstrating the GMM-FCM model's higher accuracy and reliability compared to the Spectral Clustering approach.
Conclusion: Concludes that while both models have limitations, the GMM-FCM ensemble model holds the most promise for future clinical diagnostic tools.
Keywords
Brain Tumour, MRI, Machine Learning, Unsupervised Learning, GMM-FCM, Spectral Clustering, Image Segmentation, Feature Extraction, Medical Imaging, Diagnostic Accuracy, Ensemble Model, Healthcare, Fuzzy C Means, Radiology, Anomaly Detection.
Frequently Asked Questions
What is the core focus of this research paper?
The research explores the potential of two specific unsupervised machine learning models, GMM-FCM and Spectral Clustering, to autonomously detect and classify brain tumours from MRI images without prior labelling.
What are the primary thematic areas covered?
The paper covers medical image analysis, machine learning algorithms for clustering, data preprocessing, and the evaluation of diagnostic performance in a clinical context.
What is the main objective of the study?
The objective is to determine which of the two selected unsupervised models is more accurate and better suited as a potential screening tool for identifying different types of brain tumours in clinical settings.
Which scientific methodology is utilized?
The methodology employs a comparative analysis using a standardized MRI dataset, applying feature extraction techniques, and evaluating the models through metrics like Accuracy, F1-Score, and ROC-AUC.
What topics are discussed in the main body?
The main body examines existing literature, describes the technical implementation of the clustering algorithms and feature extraction, and provides a rigorous discussion of the resulting performance data.
Which keywords define the work?
Key terms include Brain Tumour, GMM-FCM, Spectral Clustering, MRI, Unsupervised Machine Learning, and Diagnostic Accuracy.
Why did the Spectral Clustering model perform poorly in the study?
The research found that the Spectral Clustering model experienced a "catastrophic failure" in identifying "No Tumour" cases, incorrectly classifying all healthy tissue samples, which makes it currently unsuitable for clinical use.
What makes the GMM-FCM ensemble model more effective?
The GMM-FCM model provides a hybrid approach that combines probabilistic and geometric clustering, allowing it to handle mixed characteristics in tumours more effectively and resulting in a higher overall macro-average ROC-AUC compared to Spectral Clustering.
- Quote paper
- Jheirom Pablo (Author), 2026, A comparison of two machine learning models, Munich, GRIN Verlag, https://www.grin.com/document/1692018