AI in Orthopaedics 2024 Hackathon

Welcome to the repository for the AI in Orthopaedics 2024 Hackathon! This repository contains the code, datasets, and resources developed and submitted during the event.

Table of Contents

• Overview
• Features
• Installation
• Usage
• Dataset
• Team Members
• Acknowledgments

Overview

This project, developed during the AI in Orthopaedics 2024 Hackathon, uses AI to address orthopaedic challenges, focusing on detecting disc hernia via pelvic tilt, lumbar lordosis angle, sacral slope, and pelvic radius.

This project is a joint contribution from members in Southampton Emerging Therapies and Technologies (SETT) Centre, the University of Southampton, and Wessex Deanery.

Features

• Classifies disc hernia, spondylolisthesis, or normal cases.
• Custom code to handle scikit-learn Pipeline objects in downstream tasks
• Custom hierarchical model
• Interactive Predictions: A user-friendly front-end using Streamlit for predictions.
• SHAP Explanations: Provides explainable AI (XAI) outputs for model predictions using SHAP.

Installation

Follow these steps to set up the project locally:

1. Clone the repository:

   git clone https://github.com/fzhem/ai-ortho-2024.git

2. Navigate to the project directory:

   cd ai-ortho-2024

3. Install the required dependencies using either of the following methods:
   • With pip:

     pip install -r requirements.txt

   • With poetry:

     poetry install

Usage

To run the project, follow these steps:

1. The main code is in exploratory.ipynb file. This file uses custom code from extended_pipeline.py and hierarchical_model.py.
2. To run the front-end:

   streamlit run app.py

Dataset

• Source: UCI Vertebral Column
• The dataset includes 6 biomechanical attributes:
  • Pelvic Incidence
  • Pelvic Tilt
  • Lumbar Lordosis Angle
  • Sacral Slope
  • Pelvic Radius
  • Degree of Spondylolisthesis
• Labels: Normal, Hernia, Spondylolisthesis

Team Members

In alphabetical order:

• Ananya Pandey: R&D Specialist Data Analyst
• Faizan Hemotra: R&D Specialist Data Analyst
• Kehinde Makinde: R&D Specialist Data Analyst
• Lucy Bailey: Wessex Deanery Trauma and Orthopaedic Specialist Registrar
• Rory Ormiston: Wessex Deanery Trauma and Orthopaedic Specialist Registrar, University of Southampton PhD Student

Acknowledgments

We would like to thank:

• The organizers of the AI in Orthopaedics 2024 Hackathon for providing this opportunity.

---

For more details or questions, feel free to open an issue.

Licensing

• The dataset used in this project is licensed under CC BY 4.0. Proper attribution is provided in accordance with the license terms.
• The code and scripts in this repository are licensed under the MIT License. See the LICENSE file for details.