Deep Learning Approach For Brain Tumor Classification, Segmentation And Detection

• Author(s):

  Dr. K. Srinivasan | Sanjay Kumar A

• Keywords:

  Brain Tumor Detection, CNN, VGG16, Tumor Segmentation, Watershed Algorithm, Tumor Stage Prediction.

• Abstract:

  Brain Tumor Detection Remains A Critical Challenge In Medical Diagnostics. This Paper Presents A Comparative Analysis Of Classification-based, Segmentation-based, And Hybrid Deep Learning Approaches For Brain Tumor Diagnosis. The System Employs Image Processing And Convolutional Neural Networks (CNNs), Particularly The VGG16 Model, To Extract And Classify Features From MRI Scans. Tumor Stages And Regions Are Identified Using Segmentation Techniques, While Tumor Types Are Classified Using Support Vector Machines (SVM). Experimental Validation Using MRI Data From 70 Participants Revealed That The Hybrid Approach Achieved The Highest Balanced Accuracy Of 87.7%, Slightly Outperforming Classification-only (87.1%).

---

Other Details

• Paper id:

  IJSARTV11I5103611

• Published in:

  Volume: 11 Issue: 5 May 2025

• Publication Date:

  2025-05-20

Download Article