Skip to main content

Deep Learning-Assisted Brain Tumor Segmentation in MRI Imaging

Primary supervisor

Zhaolin Chen


Magnetic Resonance Imaging (MRI) stands as a cornerstone in medical imaging, providing non-invasive, high-resolution images of the human body's internal structures.  Brain tumor segmentation from MRI scans is essential for precise diagnosis and treatment planning. MRI provides detailed views of brain structures and abnormalities, but challenges like image noise, contrast imperfections and tumor variations can make segmentation difficult.

This project focusses on using advanced deep learning techniques to accurately identify tumor regions in MRI images. By leveraging convolutional neural networks (CNNs) and other machine learning algorithms such as Vision Transformers, students are required to develop a reliable framework that automates the segmentation process, improving efficiency [1,2]. 

The objective is to streamline diagnostic workflows and facilitate timely treatment decisions by reducing manual annotation process. Through collaboration between medical imaging experts, machine learning specialists, and clinicians, this project aims to advance brain tumor diagnosis and treatment, ultimately enhancing healthcare delivery.

Students involved in this project will employ advanced image processing and programming techniques, utilize cutting edge machine learning models, and gain hands-on experience with popular tools and libraries like Python, PyTorch, and TensorFlow.

Tumor segmentation network
Tumor Segmentation Network

What will students learn in this project:

Participating in this deep learning assisted brain tumor segmentation project will offer students the opportunity to:

  • Understand MRI Image Analysis: Gain insight into the analysis of Magnetic Resonance Imaging (MRI) scans, with a focus on identifying and delineating brain tumor regions.
  • Master Deep Learning Algorithms: Learn and implement advanced deep learning algorithms, particularly convolutional neural networks (CNNs), tailored for accurate segmentation of brain tumors from MRI images.
  • Practical Application of Medical Imaging: Acquire hands-on experience in applying deep learning methodologies to real-world medical imaging data, understanding their role in enhancing diagnostic accuracy.
  • Programming Proficiency: Develop proficiency in programming languages and frameworks such as Python, PyTorch, and TensorFlow, essential for implementing and optimizing deep learning models for brain tumor segmentation tasks.
  • Collaborative Research: Collaborate with experts in medical imaging and machine learning, contributing to cutting-edge research aimed at improving brain tumor diagnosis and treatment planning through automated segmentation techniques.
  • Research & Development Skills: Develop critical research abilities, learning how to design, execute, and analyze an experimental study.
  • Teamwork and Collaboration: Collaborate with a team of like-minded individuals, learning to communicate ideas effectively and working jointly towards a common goal.
  • Ethical Standards in Research: Recognize the importance of maintaining high ethical standards, particularly concerning patient data and privacy.

Student cohort

Double Semester


[1] A Robust Volumetric Transformer for Accurate 3D Tumor Segmentation,

[2] Uncertainty-guided dual-views for semi-supervised volumetric medical image segmentation,

Required knowledge

Skills Required:

Candidates interested in this project should ideally possess the following qualifications and skills:Proficiency in Programming:

  • Strong programming skills in Python, MATLAB, C/C++, or similar, necessary for developing deep learning models and image processing algorithms.Experience in Machine Learning: Prior experience or familiarity with machine learning, especially deep learning frameworks such as PyTorch and TensorFlow.
  • Problem-Solving Abilities: Strong analytical and problem-solving skills, capable of tackling complex challenges that arise during the development process.
  • Ethical Awareness: Sensitivity to the ethical considerations of working with medical data and a commitment to maintaining patient confidentiality and data security.

Application Instructions:

Candidates interested in contributing to this project should prepare an application containing the following:

  • A personal statement detailing your interest in this research project, relevant experience, and what you hope to achieve during your involvement.
  • Academic transcripts or records showcasing your previous studies and achievements.
  • Any questions or requests for further information about the project can be directed to our contact email provided below.
  • Please compile your application into a single PDF or ZIP file and submit it to our department. We look forward to your contributions to this pioneering research project.