Brain Tumour Detection Using VGG-Based Feature Extraction With Modified DarkNet-53 Model

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Machine Learning

Background:

• Malignant brain tumors pose significant health risks, necessitating early and accurate diagnosis.
• Effective treatment and improved patient outcomes depend on timely detection of brain tumors.
• Current diagnostic methods can be enhanced by advanced artificial intelligence (AI) techniques.

Purpose of the Study:

• To improve the accuracy of brain tumor detection in MRI scans using deep learning.
• To develop an AI model capable of identifying critical features in brain MRI images for early diagnosis.
• To enhance diagnostic precision for a wide range of brain tumors.

Main Methods:

• Utilized a modified DarkNet-53 deep learning architecture.
• Employed invasive weed optimization (IWO) for model optimization.
• Applied the model to a dataset of 3264 preprocessed MRI scans.

Main Results:

• Achieved a 95% success rate in brain tumor detection.
• Demonstrated superior performance compared to existing diagnostic methods.
• Successfully identified a wide range of brain tumors at an early stage.

Conclusions:

• The proposed AI-driven method significantly enhances brain tumor detection accuracy in MRI scans.
• The optimized deep learning model contributes to improved diagnostic precision and patient outcomes.
• This research highlights the potential of AI in early cancer diagnosis and treatment planning.