August 31, 2024Journal of Student Research0 citationsOpen Access

Advancing Brain Tumor Diagnosis through Machine Learning: A Comparative Study

Key Points

Convolutional Neural Networks outperformed traditional algorithms for diagnosing brain tumors.
CNN models demonstrated greater effectiveness than logistic regression and random forest in tumor classification.
Assessment compared several machine learning methods on MRI datasets featuring diverse brain tumor types.
Findings indicate that larger datasets are necessary for reliable medical application of these models.

Abstract

Brain tumor is a devastating disease affecting thousands of Americans every year. The disease requires an early and accurate diagnosis. Machine learning could be a very powerful way to speed up the diagnosis. This study explores the efficacy of various machine learning models in diagnosing and classifying brain tumors using MRI scans. Convolutional Neural Network (CNN) models were compared with traditional machine learning algorithms, including Logistic Regression, K-Nearest Neighbors (KNN), Decision Tree, and Random Forest (RF), on a dataset containing MRI images of different brain tumor types. The study came to the conclusion that the CNN was more effective than other models, and all of these models would need larger datasets before considering them usable as a medical tool.

Ask AI

Helpful

Bookmark

View Full Paper