Automatic Localization and Brand Detection of Cervical Spine Hardware

This repository contains the code for the paper titled: "Automatic localization and brand detection of cervical spine hardware on radiography using weakly supervised machine learning"

Repository Structure

• object_localization_model: Contains the code and model weights for the object localization model. The weights can be unzipped and loaded using TensorFlow
• brand_prediction_model: Contains the code and model weights for the brand identification model. The weights should be unzipped and loaded using torch.load function from PyTorch

Questions/Concerns

Please feel free to reach out to Raman Dutt or Dylan Mendonca if you have any questions on the implementations within this repository