Brain Tumor Segmentation Using Convolutional Neural Networks in MRI Images.

Area of Science:

• Medical Image Processing
• Artificial Intelligence in Medicine
• Neurology

Background:

• Brain tumor segmentation is crucial for early detection and treatment, significantly impacting patient survival.
• Manual segmentation is time-consuming and prone to errors, necessitating automated solutions.
• Existing automated methods may lack precision and efficiency in complex cases.

Purpose of the Study:

• To develop an optimized automatic brain tumor segmentation method using Enhanced Convolutional Neural Networks (ECNN).
• To improve the accuracy and efficiency of Magnetic Resonance Imaging (MRI) based tumor segmentation.
• To leverage BAT algorithm for loss function optimization in deep learning models for medical imaging.

Main Methods:

• Proposed an Enhanced Convolutional Neural Network (ECNN) architecture with small kernels to mitigate overfitting.
• Utilized a BAT algorithm for optimizing the loss function of the segmentation model.
• Implemented pre-processing steps including skull stripping and image enhancement algorithms.
• Applied the optimized ECNN model for automatic segmentation of brain tumors in MRI scans.

Main Results:

• The proposed ECNN method demonstrated superior performance in brain tumor segmentation compared to existing techniques.
• Evaluated performance metrics included precision, recall, and accuracy, showing significant improvements.
• The optimization strategy effectively enhanced the segmentation accuracy and reliability.

Conclusions:

• The BAT algorithm-optimized ECNN provides an effective and efficient solution for automatic brain tumor segmentation in MRI.
• This automated approach holds potential for improving diagnostic speed and accuracy in clinical practice.
• Future work may explore advanced selection schemes to further enhance segmentation accuracy.