Inspiration
A personal hairline fracture injury that was misdiagnosed by doctors
What it does
It identifies hairline fractures in medical imaging
How we built it
We preprocessed the data, then trained and validated the model, then tested the model
Challenges we ran into
We initially had an unequal weighting for the two classes that we trained the model on (fractured vs. non-fractured), but then fixed it by creating an equal amount of data in each class
Accomplishments that we're proud of
We're proud of our model's accuracy: 96%!
What we learned
We learned more about convoluted neural networks
What's next for OsteNet: A Novel Tool For Hairline Fractures In Bone Disease
Real-world application in the medical field, incorporation into an application or product
*data is from kaggle.com (links within notebook)
Built With
- colab
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.