MRI Brain Tumor Image Classification Using a Combined Feature and Image-Based Classifier.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• Accurate brain tumor classification is crucial for patient prognosis and treatment planning.
• Existing methods often rely on single feature or image types, limiting classification performance.

Purpose of the Study:

• To develop and evaluate a novel Combined Feature and Image-based Classifier (CFIC) for brain tumor classification.
• To compare the performance of CFIC against various deep neural network and deep convolutional neural network-based classifiers.

Main Methods:

• Utilized deep neural network (DNN) and deep convolutional neural network (DCNN) architectures.
• Proposed several classifiers including actual image feature-based (AIFC), segmented image feature-based (SIFC), and combined approaches (CFIC).
• Trained and tested classifiers on the Kaggle Brain Tumor Detection 2020 dataset.

Main Results:

• The proposed CFIC demonstrated superior performance compared to all other evaluated methods.
• CFIC achieved high sensitivity (98.86%), specificity (97.14%), and accuracy (98.97%).
• Results showed significant improvements over existing brain tumor classification techniques.

Conclusions:

• The CFIC approach offers a robust and highly accurate method for brain tumor classification.
• Integrating both feature and image data enhances diagnostic capabilities.
• This method holds promise for improving clinical decision-making in neuro-oncology.