MedAI aims to automate surgical instrument recognition using a camera and an AI model.
- Create a dataset of surgical instrument images.
- Develop a machine learning model for instrument classification.
- Deploy the model for production use.
- Automate tasks
- Improved efficiency
- Reduced human errors
- Achieving 100% accuracy is challenging.
- The model's initial capabilities will be limited to instruments in the training dataset.
- Performance may be affected by image quality, lighting, and irrelevant images.
- The model may need updates for new instruments or different environments.
- I made a demo where you can use the model to recognize 5 different surgical instruments here.
- I've trained several models in TensorFlow, PyTorch, and YOLOv5. For more info, check my code.