Breast cancer diagnosis using a CNN. Used CBIS-DDSM breast cancer dataset. Model is developed, trained, and hosted with Google Cloud Platform.
What things you need to install the software and how to install them
Google Cloud
Keras and the pre-reqs
Tensorflow and the pre-reqs
Numpy
NBIA Data Retriever (For downloading files from official database)
Sign up for an GCP account Login in GCP with gcloud init to setup environment Uncomment the first line in gcloud.run.sh or local_gcloud_run.sh to save login credentials locally
Three classes: Benign, Benign without callback, Malignant
Class Distrubution: Train: Benign:1105, Benign without callback:578, Malignant:1181, Total: 2864 Test: Benign:324, Benign without callback:104, Malignant:276, Total: 704
Configure config.yaml for specific Google ML Engine instance Run gcloud.run.sh
- Google Cloud Platform - The cloud platform used
- Google Cloud ML Engine - Service used to train and test the CNN model
- Google Cloud App Engine - Service used to host the REST API (future plans)
- Brian Nguyen - Initial work - briannoogin
This project is licensed under the MIT License - see the LICENSE.md file for details