Early Diagnosis of Brain Tumour MRI Images Using Hybrid Techniques between Deep and Machine Learning

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Oncology

Background:

• Brain tumors are a destructive disease, and accurate diagnosis is vital for effective treatment and improved patient survival.
• Misdiagnosed brain tumors can lead to detrimental interventions, reducing survival chances.
• Computer-aided diagnostic systems, particularly those using deep and machine learning, show promise in enhancing diagnostic accuracy.

Purpose of the Study:

• To evaluate the efficacy of hybrid deep learning and traditional machine learning techniques for accurate brain tumor diagnosis.
• To compare the performance of different deep learning models (AlexNet, ResNet-18) combined with machine learning classifiers (SVM, SoftMax) for brain tumor classification.

Main Methods:

• Brain tumor magnetic resonance imaging (MRI) images were preprocessed using an average filter.
• Deep learning models (AlexNet, ResNet-18) were employed to extract features using deep convolutional layers.
• Extracted features were classified using Support Vector Machine (SVM) and SoftMax classifiers.

Main Results:

• The hybrid approach combining deep learning feature extraction with machine learning classification demonstrated superior performance.
• The AlexNet+SVM hybrid technique achieved the highest accuracy (95.10%), sensitivity (95.25%), and specificity (98.50%).
• All tested systems showed significant results in classifying brain tumors from MRI scans.

Conclusions:

• Hybrid deep learning and machine learning models offer a powerful tool for accurate and early brain tumor diagnosis.
• The AlexNet+SVM combination presents a highly effective method for brain tumor classification, improving diagnostic capabilities.