BrainScan Tumour Detector

M-RCNN input vs. output (reference)
intel Cloud Developer env
User Interface
Mask R-CNN model training...
Mask R-CNN model troubleshooting... (1/2)
Mask R-CNN model troubleshooting... (2/2)

(Team 3)

Inspiration

https://www.kaggle.com/code/ruslankl/brain-tumor-detection-v2-0-mask-r-cnn/notebook

What it does

Using a Mask R-CNN model, it analyses medical images, providing accurate and rapid identification of brain tumours. Simplifying diagnostic precision and empowering healthcare professionals to make timely decisions.

How we built it

Python (Jupyter Notebook), TensorFlow, Tkinter, intel Developer Cloud, SQL
We built the brain tumour detection software using Python and the MVC (Model-View-Controller) architectural pattern. The model leverages SQL for data storage, and the GUI is developed with Tkinter. -- The system allows for image upload, patient data management, and comment logging, providing a comprehensive tool for medical professionals.

Challenges we ran into

Mask R-CNN model: initially built for tf1.x, had to be rebuilt for python3.x and depending modules.
Connecting the SQL server to the model posed challenges in terms of establishing a seamless integration. The process required troubleshooting to achieve a robust connection between the SQL server and the machine learning model.

Accomplishments that we're proud of

2nd hackathon!
In the process, we achieved valuable accomplishments by delving into cutting-edge technologies. This venture provided us with significant exposure to new tools, frameworks, and methodologies, enhancing our skill set and broadening our technological horizons

What we learned

Cloud development, mrcnn, Tkinter, GUI, …
Throughout the project, we gained profound insights into SQL databases and the Model-View-Controller (MVC) architecture. We navigated the intricacies of database management, honed our SQL skills to efficiently store and retrieve data, and implemented the MVC pattern to structure our application for enhanced modularity and maintainability

What's next for BrainScan Tumour Detector

Keep working on the project!
In the next phase, we aim to elevate the user experience by implementing additional software features. Our focus will be on enhancing the UI to ensure efficiency and user-friendly interactions. We plan to incorporate advanced functionalities, streamline processes, and optimize the overall performance to provide a seamless and intuitive experience for our users.

Built With

Submitted to

HackNotts '84
- Winner intel - 2nd, 3rd

Created by

I worked on the Mask R-CNN backend, dealing with image segmentation. It was my first time with Jupyter Notebook and intel Cloud Developer: I manipulated image data, refined model components and optimised memory usage.

Nathan Sidi Bakari
'sup, I’m Nabakator. I’m interested in space, coding and science. I’m currently learning Aerospace Engineering @ UON
I worked on developing the software to display our model including front end and back end with Python using MVC as well as integrating SQL to manage patient's data.

Angelo Rosu