Related Experiment Video
Updated: Jul 16, 2025
09:53
Quantifying the Brain Metastatic Tumor Micro-Environment using an Organ-On-A Chip 3D Model, Machine Learning, and Confocal Tomography
Published on: August 16, 2020
7.3K
Explainable Convolutional Neural Networks for Brain Cancer Detection and Localisation.
Francesco Mercaldo1,2, Luca Brunese1, Fabio Martinelli2
1Department of Medicine and Health Sciences "Vincenzo Tiberio", University of Molise, 86100 Campobasso, Italy.
Sensors (Basel, Switzerland)
|September 9, 2023
Summary
This study introduces a deep learning method for detecting and localizing brain cancer using magnetic resonance images. The approach achieved high accuracy, offering explainable AI insights for improved cancer diagnosis.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Brain cancer is a highly aggressive malignancy with a poor survival rate.
- Accurate and early detection of brain tumors is crucial for patient outcomes.
- Current diagnostic methods can be enhanced with advanced computational techniques.
Purpose of the Study:
- To develop and evaluate a deep learning-based method for detecting and localizing brain cancer from MRI scans.
- To enhance the explainability of AI models in medical image analysis.
- To assess the performance of various convolutional neural network architectures for this task.
Main Methods:
- Utilized deep learning, specifically convolutional neural networks (CNNs) and class activation mapping (CAM).
- Analyzed a dataset of 3000 magnetic resonance images (MRIs).
- Evaluated four distinct CNN models: VGG16, ResNet50, AlexNet, and MobileNet.
Main Results:
- Achieved high accuracy rates for brain cancer detection, ranging from 97.83% to 99.67%.
- Class activation mapping provided visual explanations, highlighting image regions indicative of cancer.
- Demonstrated the effectiveness of multiple CNN architectures in identifying brain tumors.
Conclusions:
- The proposed deep learning method is effective for accurate brain cancer detection and localization using MRI.
- Explainable AI techniques improve the interpretability of diagnostic models.
- The findings support the integration of advanced AI in neuro-oncology for improved diagnostic capabilities.
