A Comprehensive Machine Learning Framework for Early Brain Tumor Diagnosis from MRI Data

View at publisher Summarise with VIRAT
Authors: Veerasekar N, Malani M, Narmadha S, Thrisha S, Allwin Felix S
Cited by: 0
References: 12

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

In brain tumor, early and precise detection is still a major challenge because diagnosis has an immediate effect on treatment results and patient survival. A feature-engineered machine learning framework for recognizing brain tumor from MRI data is demonstrated in this work. A comprehensive feature space that includes statistical, textural (based on GLCM), and advanced image quality metrics like PSNR, SSIM, MSE, and DC has been generated using two Kaggle datasets. The classification performance of three machine learning algorithms- Logistic Regression, XG Boost, and Random Forest-was analyzed. The Random Forest model had the best accuracy (0.9909), F1-score (0.9948), and perfect recall (1.0000) out of all of them, indicating that it had no false negatives. The most important factor in discerning between malignant and non-malignant tissues, according to feature importance analysis, was image quality metrics. These results demonstrate how interpretable, non-invasive machine learning models can be used as reliable tools for bettering brain tumor diagnostic precision.

Fields of Study & Subjects

Brain Tumor Detection and Classification AI in cancer detection Radiomics and Machine Learning in Medical Imaging

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
1141–1146
Publication date
21 January 2026
Language
EN