Abstract:Detecting brain tumors in their early stages is crucial. Brain tumors are classified by biopsy, which can only be performed through definitive brain surgery. Computational intelligence-oriented techniques can help physicians identify and classify brain tumors. Herein, we proposed two deep learning methods and several machine learning approaches for diagnosing three types of tumor, i.e., glioma, meningioma, and pituitary gland tumors, as well as healthy brains without tumors, using magnetic resonance brain images to enable physicians to detect with high accuracy tumors in early stages.

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve This paper aims to address the issue of early detection of brain tumors. Brain tumors are a serious condition that can lead to death if not treated promptly. Currently, the diagnosis of brain tumors often requires a biopsy through surgery, which is both complex and carries certain risks. Therefore, researchers hope to develop a non-invasive method to improve the accuracy of early detection of brain tumors. Specifically, this paper proposes a deep learning and machine learning method based on Magnetic Resonance Imaging (MRI) to identify and classify three types of brain tumors (glioma, meningioma, and pituitary adenoma), as well as healthy brains without tumors. Through these methods, doctors can detect brain tumors with high accuracy at an early stage, thereby improving patient survival rates. ### Main Contributions 1. **Dataset**: A dataset containing 3264 T1-weighted contrast-enhanced MRI images was used, covering gliomas, meningiomas, pituitary adenomas, and healthy brains. 2. **Network Architecture**: The internal structure of 2D Convolutional Neural Networks (CNN) and convolutional autoencoder networks was designed and adjusted, including the number of layers, the arrangement of layers, the types of parameters, and their values, to optimize the model's accuracy. 3. **Feature Extraction and Classification**: Using the extracted key features, the classification of three types of brain tumors and healthy brains was performed through 2D CNN, autoencoder networks, and six common machine learning techniques. 4. **Performance Evaluation**: The proposed 2D CNN and autoencoder networks achieved training accuracies of 96.47% and 95.63%, respectively, with average recall rates of 95% and 94%, and areas under the ROC curve of 0.99 or 1. 5. **Statistical Analysis**: Through one-way ANOVA of accuracy, recall, and F-values, significant differences (p-value < 0.001) were shown between the proposed methods and other machine learning methods. 6. **Generalization Ability**: The proposed architecture outperformed other state-of-the-art methods on the MRI dataset and demonstrated higher generalization ability. ### Related Work This paper also reviews recent research on brain tumor classification based on MRI images, highlighting the advantages and limitations of these studies and proposing improvements based on them. For example, some studies have achieved high classification accuracy but lack performance comparison with other traditional machine learning methods; other studies have issues with computational complexity. By comprehensively considering these issues, this paper proposes a more efficient and accurate solution.

MRI-based brain tumor detection using convolutional deep learning methods and chosen machine learning techniques

Early Diagnosis of Brain Tumour MRI Images Using Hybrid Techniques between Deep and Machine Learning

Ensemble deep learning for brain tumor detection

Convolutional Block Attention Module-based Deep Learning Model for MRI Brain Tumor Identification (ResNet-CBAM)

Brain Tumor Detection Using Deep Learning Approaches

Deep Learning-Based Segmentation and Classification Techniques for Brain Tumor MRI: A Review

Machine Learning Approach to Brain Tumor Detection and Classification

Efficient Feature Extraction and Classification Architecture for MRI-Based Brain Tumor Detection

Image-based MRI detection of brain tumours using convolutional neural networks

Deep Learning for Smart Healthcare-A Survey on Brain Tumor Detection from Medical Imaging

Brain tumor detection from MRI images using deep learning techniques

Predictive modeling of brain tumor: A Deep learning approach

Efficient Brain Tumor Detection Based on Deep Learning Models

Classification of MRI Brain Tumor Images using Deep Learning Segment Anything Model for segmentation and Deep Convolution Neural Network

A robust MRI-based brain tumor classification via a hybrid deep learning technique

A review on detecting brain tumors using deep learning and magnetic resonance images

Brain Tumor Detection using Convolution Neural Network

A survey of deep learning for MRI brain tumor segmentation methods: Trends, challenges, and future directions

A Deep Analysis of Brain Tumor Detection from MR Images Using Deep Learning Networks

Enhancing brain tumor classification in MRI scans with a multi-layer customized convolutional neural network approach

Study on Brain Tumor Classification Through MRI Images Using a Deep Convolutional Neural Network