Spinal-QDCNN: advanced feature extraction for brain tumor detection using MRI images.

Area of Science:

• Medical imaging analysis
• Artificial intelligence in healthcare
• Neuro-oncology

Background:

• Brain tumors pose significant health risks, necessitating early detection for improved patient outcomes.
• Current brain tumor detection methods lack sufficient accuracy and struggle with subtle cellular changes.
• Existing models often exhibit inefficient learning architectures, limiting their diagnostic capabilities.

Purpose of the Study:

• To introduce an advanced model for accurate brain tumor detection using MRI images.
• To overcome the limitations of existing methods in detecting subtle changes in brain cells.
• To enhance the diagnostic accuracy and efficiency of brain tumor identification.

Main Methods:

• A novel SpinalNet-Quantum Dilated Convolutional Neural Network (Spinal-QDCNN) model was developed.
• The methodology involved MRI image pre-processing, including RoI extraction and thresholding-based enhancement.
• Image segmentation utilized Projective Adversarial Networks (PAN), followed by augmentation and extensive feature extraction (statistical, Gabor, DWT, GBP).

Main Results:

• The proposed Spinal-QDCNN model achieved a maximum accuracy of 86.356%.
• The model demonstrated high sensitivity (87.37%) and specificity (88.357%) in brain tumor detection.
• The integration of QDCNN and SpinalNet proved effective for enhanced diagnostic performance.

Conclusions:

• The Spinal-QDCNN model offers a promising advancement in brain tumor detection from MRI data.
• The method effectively addresses limitations of traditional approaches, improving detection of subtle abnormalities.
• The achieved accuracy, sensitivity, and specificity highlight the potential of this model for clinical application.