🧠 Brain Tumor MRI Classification (CS231: Introduction to CV)

---

🌟 Overview

This project delves into Brain Tumor MRI Classification, providing our initial foray into research-like endeavors. It aims to enhance skills in literature review, solution mapping, and applying techniques introduced in our course. We implement a standard Image Classification pipeline and explore creative, data-driven improvements by analyzing insights and errors. The project concludes by comparing our solutions to related work, drawing valuable conclusions, and identifying future work opportunities.

---

👥 Used By

This project serves as a resource for:

• University Students: Learning how to undertake final research projects.
• Researchers: Interested in brain tumors or the medical field.
• Guidance Seekers: Those looking to understand and code using CV frameworks and online GPU platforms.

---

Roadmap

1. Literature Review and Solution Mapping

   • Identify feasible topics and refine them to narrow scopes.
   • Conduct research to understand pipeline, techniques, and direction mapping.
     • Search some papers in the field or relation to have a general understanding of pipeline, techniques, and direction mapping.

2. Image Classification Pipeline

   • Image Pre-processing: Resize, format channel form, and adjust pixel value range.
   • Image Enhancement
   • Feature Extraction
   • Feature Reduction: PCA
   • Training Classifier Model
   • Evaluating Model

3. Insight and Error Diagnosis

   • Propose novel methods and analyze reasons for proposals.

4. Project Presentation and Documentation

   • Design a formal PowerPoint presentation.
   • Record research process and results using Google Docs.

🛠 Tech Stack

• Process Record: Google Docs
• Presentation: PowerPoint
• Topic Finding: Jenni-AI, Google Search, Gemini
  • Find driven-dataset > topic > problem. Use Jenni-AI and Google Search to find papers, and Gemini to explain them and recommend (link them) some trend or value directions.
• Frameworks & Libraries:
  • Image Operations: numpy, skimage
  • Dataset Exploration & Display: pandas, matplotlib, skimage, opencv
  • Feature Extraction & Reduction: numpy, skimage, sklearn
  • Dataset Splitting: sklearn (train_test_split)
  • PCA: sklearn (pca)
  • Training & Evaluation: sklearn, keras, sklearn-cross_validation

---

📘 Lessons Learned

• Approaching a first research project.
• Understanding image ML workflow in research.
• Gaining proficiency in CV frameworks.

---

📢 Feedback

For any feedback or inquiries, please contact us at 22521178@gm.uit.edu.vn.

📜 License

This project is licensed under the MIT License.

---