Brain Tumor Classification Using MobileNetV2 and SVM
Author Information
Author(s): Adamu Mohammed Jajere, Kawuwa Halima Bello, Qiang Li, Nyatega Charles Okanda, Younis Ayesha, Fahad Muhammad, Dauya Salisu Samaila, Giménez-Llort Lydia, Battaglia Simone, Tanaka Masaru, Roesler Rafael, Fred Lam C.
Primary Institution: Tianjin University
Hypothesis
Can a hybrid model combining MobileNetV2 and Support Vector Machine improve brain tumor classification accuracy and efficiency?
Conclusion
The MobileNetV2-SVM hybrid model shows high accuracy and efficiency in classifying brain tumors from MRI images.
Supporting Evidence
- The model achieved AUC scores of 0.99 for glioma, 0.97 for meningioma, and 1.0 for pituitary tumors.
- The hybrid model improves classification accuracy while reducing computational overhead.
- MobileNetV2's lightweight architecture makes it suitable for deployment in resource-constrained environments.
- The model can assist in quicker and more accurate diagnoses of brain tumors.
Takeaway
This study created a smart computer program that helps doctors tell if someone has a brain tumor by looking at pictures of their brain.
Methodology
The study used a hybrid model that combines MobileNetV2 for feature extraction and SVM for classification, trained on a dataset of MRI images.
Limitations
The model's complexity may hinder rapid real-time diagnosis in certain clinical environments.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website