Evaluating Brain Tumor Detection with Deep Learning Convolutional Neural Networks Across Multiple MRI Modalities.

Area of Science:

• Neuroimaging
• Artificial Intelligence
• Oncology

Background:

• Central Nervous System (CNS) tumors pose significant health risks.
• Magnetic Resonance Imaging (MRI) is vital for brain tumor detection and diagnosis.
• Deep learning, especially Convolutional Neural Networks (CNNs), offers potential for improved diagnostic accuracy.

Purpose of the Study:

• To evaluate the diagnostic performance of MRI sequences for brain tumor detection using CNNs.
• To assess the effectiveness of transfer learning techniques in enhancing CNN architectures for brain tumor identification.
• To identify optimal combinations of MRI sequences and CNN models for clinical screening.

Main Methods:

• Utilized six fundamental MRI sequences across four distinct CNN architectures.
• Employed transfer learning techniques to enhance CNN performance.
• Analyzed a dataset of 1646 MRI slices from 62 patients, including tumor-bearing and normal cases.

Main Results:

• Achieved a classification accuracy of 98.6% in detecting tumor-involved brain slices.
• Evaluated the performance of each MRI sequence across different CNN models.
• Identified specific MRI sequences and CNN combinations that are highly effective for screening.

Conclusions:

• CNN-based deep learning models demonstrate high potential for accurate brain tumor detection using MRI.
• The findings provide insights into integrating AI with MRI for enhanced diagnostic workflows.
• Optimized combinations of MRI sequences and CNNs can effectively support radiologists in screening for brain tumors.