SMART SCAN: HYBRID DEEP LEARNING AND MACHINE LEARNING FRAMEWORK FOR MRI-BASED BRAIN TUMOR DETECTION

Abstract

Brain tumor detection and classification from MRI scans play a critical role in early diagnosis and effective treatment planning. However, manual interpretation of medical images is time-consuming, error-prone, and highly dependent on expert knowledge. This study introduces Smart Scan, an intelligent hybrid framework that combines Convolutional Neural Networks (CNNs) for deep feature extraction with traditional Machine Learning (ML) classifiers for accurate tumor identification. The proposed system processes raw MRI images through a pre-trained or custom CNN architecture to extract spatial and texture-based features, which are then fed into selected ML models such as Support Vector Machines (SVM), Random Forest (RF), and k-Nearest Neighbors (k-NN) for final classification. Extensive experiments were conducted using benchmark MRI brain tumor datasets, and the hybrid model demonstrated superior accuracy, precision, and sensitivity compared to standalone deep learning or ML approaches. The results affirm that Smart Scan effectively leverages the strengths of both learning paradigms, offering a robust and scalable solution for real-time, non-invasive brain tumor diagnosis.