Based on patient's brain MRI detect whether he has or not brain tumor and if he has - produce location of this tumor
This project is my 1st experience with Image segmentation problem. I hope that I did good job.
In this task we're going to build 2 models and solve 2 problems:
The 1st one will be responsible for classification (whether patient has brain tumor or not). This model is build via transfer learning, exactly ResNet50 with Fine tuning. The results are impressive (accuracy 0.989, recall 0.99 and precission 0.98).
The 2nd model will detect location of brain tumor for those patients who has it. This model is build via ResUnet architecture, uses Tversky loss function and metric. Results of the model were great: all rates have +-99% (including recall and precision).
In the end you can see final visualisations of model's results and useful links.
Brain MRIs and ready model architectures with weights you can download from here: https://drive.google.com/drive/folders/1v-WpOl6Bfj4LY-L9-D5mjUrzPfo-ivlt?usp=sharing
This project is a part of course on Udemy.com.
-
Notifications
You must be signed in to change notification settings - Fork 0
mariaKozlovtseva/Brain-tumor-via-MRI-detection
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Based on patient's brain MRI detect whether he has or not brain tumor and if he has - produce location of this tumor
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published