ChatSkD (Chat GPT based Skeletal Dysplasia aware bot)

Overview

ChatSkD is a context-aware GPT-4 based chatbot designed to provide differential diagnoses in skeletal dysplasia, aiming to enhance AI-assisted diagnostic tools in the field.

Highlights

Demonstrates high performance in Skeletal Dysplasia diagnosis.
Outperforms generic GPT-4 models in providing contextual responses and diagnostic support.
Potential Impact: Could significantly improve the efficiency and accuracy of diagnostic processes in radiology.

Important Note

Status of Research: This project is part of an ongoing research initiative and the details of the study may change pending peer review.
Caution for Use: The tool is not intended for medical use and should be used accordingly.

Flask Application for Query Processing

This Flask application is designed to process queries using a GPT-4 based model, extracting relevant information from a specified document directory and serving it through a web interface.

Prerequisites

Before you begin, ensure you have met the following requirements:

Python 3.x installed on your system.
Flask and other dependencies installed (use pip install -r requirements.txt).
An OpenAI API key.

Installation and Setup

Clone the Repository:

Use the following command to clone the giaCB repository:

git clone https://github.com/aaekay/ChatSkD

Then, change to the repository directory:

cd ChatSkD

Install Dependencies:

Instal Minconda first

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

Now activate conda:

source ~/.bashrc

Now use the environment file to create envrionment skd

conda env create -f environment.yml

If above code throws error then do the following

conda create -n skd python=3.8
conda activate skd
pip install -r requirements.txt

Set OpenAI API Key:

You need to set up the API key as follows

touch .env

Now open the file and paste the key as follows OPENAI_API_KEY=

Running the Index Creation

Prepare Your Environment:

In the above setps you created your env

Specify Folder Paths:

The script requires two command-line arguments:
pdf_folder: The path to the folder containing the PDF documents.
data_folder: The path to the folder where the indexed data will be saved.

Run the Script:

Execute the index creation script using the following command:

python skd_data_processing.py [path_to_pdf_folder] [path_to_data_folder]

Replace [path_to_pdf_folder] and [path_to_data_folder] with the respective folder paths.

Verification:

After successful execution, the script will process the documents in the specified PDF folder and save the indexed data in the index folder.
A confirmation message will be displayed: "Documents processed and saved to [data_folder]."

Running the Application

Start the Flask Server:

Run the Flask application using the following command:

python skd_app.py [index_folder]

Optionally, specify the index folder path as a command-line argument. If not provided, the application will prompt for it.

Accessing the Application:

Navigate to http://localhost:8000 (or the appropriate port) in your web browser.

Using the Application:

Enter a query in the provided text box and submit.
The application will process the query and display the results, including relevant document references.

License

This project is licensed under the MIT Licence. See the LICENSE file for more details.

aaekay/ChatSkD