app.py

import os
import base64
import zipfile
from pathlib import Path
import streamlit as st
from byaldi import RAGMultiModalModel
from openai import OpenAI

# Function to unzip a folder if it does not exist
def unzip_folder_if_not_exist(zip_path, extract_to):
    if not os.path.exists(extract_to):
        with zipfile.ZipFile(zip_path, 'r') as zip_ref:
            zip_ref.extractall(extract_to)

# Example usage
zip_path = 'medical_index.zip'
extract_to = 'medical_index'
unzip_folder_if_not_exist(zip_path, extract_to)

# Preload the RAGMultiModalModel
@st.cache_resource
def load_model():
    return RAGMultiModalModel.from_index("medical_index")

RAG = load_model()

# OpenAI API key from environment
api_key = os.getenv("OPENAI_API_KEY")
client = OpenAI(api_key=api_key)

# Streamlit UI
st.title("Medical Diagnostic Assistant")
st.write("Enter a medical query and get diagnostic recommendations along with visual references.")

# User input
query = st.text_input("Query", "What should be the appropriate diagnostic test for peptic ulcer?")

if st.button("Submit"):
    if query:
        # Search using RAG model
        with st.spinner('Retrieving information...'):
            try:
                returned_page = RAG.search(query, k=1)[0].base64

                # Decode and display the retrieved image
                image_bytes = base64.b64decode(returned_page)
                filename = 'retrieved_image.jpg'
                with open(filename, 'wb') as f:
                    f.write(image_bytes)

                # Display image in Streamlit
                st.image(filename, caption="Reference Image", use_column_width=True)

                # Get model response
                response = client.chat.completions.create(
                    model="gpt-4o-mini-2024-07-18",
                    messages=[
                        {"role": "system", "content": "You are a helpful assistant. You only answer the question based on the provided image"},
                        {
                            "role": "user",
                            "content": [
                                {"type": "text", "text": query},
                                {
                                    "type": "image_url",
                                    "image_url": {"url": f"data:image/jpeg;base64,{returned_page}"},
                                },
                            ],
                        },
                    ],
                    max_tokens=300,
                )
                
                # Display the response
                st.success("Model Response:")
                st.write(response.choices[0].message.content)
            except Exception as e:
                st.error(f"An error occurred: {e}")
    else:
        st.warning("Please enter a query.")