SwingAI

This is the beginning of the SwingAI Project!

Conceptual overview

This project aims to develop an AI-powered web application to assist in tracking and analyzing Unilateral Cerebral Palsy (UCP) in pediatric patients before and after botox treatment. The application provides a more advanced and objective method for assessing UCP conditions compared to traditional physical marker-based approaches.

User Roles

Doctors/Medical Professionals:

• Upload patient videos
• View AI-generated analysis of arm movement
• Access graphs of patient conditions

Authorized Hospital Staff:

• Log in through secure hospital authentication system
• Manage patient records and video uploads

Functional Requirements

Login Page

• Secure authentication through hospital's system
• Restricted access for authorized personnel only

Homepage

• Simple interface
• Button to navigate to video upload page

Video Upload Page

• Upload up to 2 patient videos
• Support for video file processing
• Display processing animation during analysis
• Option of using different AI models for analysis
• Accurate AI analysis

Results Page

• Display AI-generated analysis
• Graphical representation of arm movement angles
• Detailed insights into patient's UCP condition

Tech Stack

Web Application

• Frontend Framework: Next.js
• Backend: Next.js (integrated backend)
• Authentication: Magic Link (Temporary)
  • Hospital's authentication system (to be implemented)

AI Model (Not implemented yet)

• Video Tracking: MediaPipe
• ML models: SciKit-Learn
• Graphing: Numpy
• Other ML/Data Analytic tools: Pandas

Cloud (Not implemented yet)

• Cloud Provider: Amazon Web Services (AWS)
• Storage: S3 Bucket
• Queuing: SQS
• Compute: Lambda Functions

Migration Scripts

• No existing data migration required at this stage

Development Environment Setup

Prerequisites

• Ensure Node.js is installed
• Ensure Python 3.12 is installed
• Docker (Optional, not implemented yet)
• AWS CLI configured with appropriate credentials (Optional, not implemented yet)

Database Initialization

• No local database setup required (no database implemented yet)

Authentication Setup

• Coordinate with hospital IT for authentication integration
• Implement Auth0 or similar service in future iterations

Starting Project

Running Front-End

1. Clone the repository:

git clone https://github.com/UTDallasEPICS/Swing.git

2. Install dependencies:

# npm
npm install

# pnpm
pnpm install

# yarn
yarn install

# bun
bun install

4. Run the development server

# npm
npm run dev

# pnpm
pnpm run dev

# yarn
yarn dev

# bun
bun run dev

Running ML currently

1. Clone the repository:

git clone https://github.com/UTDallasEPICS/Swing.git

2. Change directory to ml folder

cd ml

3. Run AI_Method.py or other .py files in the folder

# Running AI_Method.py
Python AI_Method.py

# General template
Python [pythonfile]

Future Instructions (If AWS is implemented)

AWS configurations

• Use AWS Managment Console
• Login as root email/user
• Need MFA which is already configured
• Go into "Users"
• Click create "Access Key" for user (This creates access keys for users)