Related Experiment Video
Updated: Nov 19, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
An efficient approach to diagnose brain tumors through deep CNN
Bakhtyar Ahmed Mohammed1, Muzhir Shaban Al-Ani2
1University of Human Development, College of Science and Technology, Department of Computer Science, Sulaymaniyah, KRG, Iraq and University of Sulaimani, College of Science, Department of Computer, Sulaymaniyah, KRG, Iraq.
This study developed an accurate deep convolutional neural network (CNN) method for brain tumor diagnosis using MRI images, achieving 96% accuracy. The deep CNN approach enhances early detection and treatment success rates for various brain tumor types.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Brain tumors pose significant health risks, necessitating early and accurate detection for effective patient treatment.
- Deep convolutional neural networks (CNNs) show promise for analyzing medical images, particularly MRI scans, for tumor identification.
Purpose of the Study:
- To propose and evaluate a systematic deep CNN-based approach for diagnosing brain tumors using MRI images.
- To assess the accuracy, sensitivity, and error rates of the proposed deep CNN model.
Main Methods:
- A dataset of 1258 MRI images (4 classes: ependymoma, meningioma, medulloblastoma, normal brain) was curated from the Radiopedia database.
- Deep CNNs were employed for feature learning and classification, with data split into training (70%), testing (20%), and validation (10%) sets.
- Image processing and analysis were performed using MATLAB software.
Main Results:
- The deep CNN approach achieved an overall accuracy of 96% after 15 epochs.
- Feature learning accuracy and sensitivity increased from 47.02% (1 epoch) to 96% (15 epochs).
- The error rate decreased from 52.98% to 4% with an increase in epochs from 1 to 15.
Conclusions:
- Deep CNNs offer an efficient method for feature learning, extraction, and classification of brain tumors from MRI images.
- The proposed deep CNN model demonstrates high accuracy (96%) and effectiveness in brain tumor diagnosis.
- Optimizing the number of epochs (e.g., to 15) is crucial for maximizing the accuracy and sensitivity of the deep CNN approach.
