Automated surgical tool detection and segmentation of surgical images application using Deep-Learning
The surgical procedure and instruments are detected using two deep learning models, namely UNet and Yolov5 respectively.Used YOLOv5 for Object detection and U-net architecture for segmentation tasks.
The front-end of the application is built using React JS, which accepts the surgical image and displays the results of the deep learning models. The user can view the instrument for uploaded images.
MongoDB is used as the database to store the Authentication. Node JS is used as the backend framework to handle the communication between the front-end and the database.
Before installing the project, ensure that you have the following software installed on your machine:
- Node js v16
- git
- VS code
- Clone the project repository:
git clone https://github.com/AutoSurgery/AutoSurgery.git- Navigate to the project directory:
cd AutoSurgery-main- Install the dependencies for the backend:
npm install- Start the backend server:
npm start- In a new terminal window, navigate to the project directory and install the dependencies for the frontend:
cd client
npm install- Start the frontend server:
npm start- Open your web browser and go to
http://localhost:3000to view the application.
Note: Make sure that MongoDB url is right is running before starting the backend server and is connected to database
Now you can register yourself if you are not resgistered
.Once you login you can upload your images here
Choose File and click on PREDICT IMAGE
Make sure your Flask Server is turned on
You can now view the predicted tools and the segmented mask and labels
Contributions are welcome and appreciated. To contribute, follow these steps:
- Fork the repository to your own GitHub account.
- Clone your forked repository to your local machine.
- Create a new branch for your changes.
- Make your changes and commit them to your branch.
- Push your changes to your forked repository.
- Open a pull request to the main repository.
This project is licensed under the MIT License.
If you have any questions or comments about the project, please feel free to contact us