SURESHBEEKHANI's picture
Upload 2 files
667850f verified
raw
history blame
8.57 kB
import streamlit as st
from PIL import Image
import os
import base64
import io
from dotenv import load_dotenv
from groq import Groq
from reportlab.lib.pagesizes import letter
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer
from reportlab.lib.styles import getSampleStyleSheet
# ======================
# CONFIGURATION SETTINGS
# ======================
PAGE_CONFIG = {
"page_title": "Radiology Analyzer",
"page_icon": "🩺",
"layout": "wide",
"initial_sidebar_state": "expanded"
}
ALLOWED_FILE_TYPES = ['png', 'jpg', 'jpeg']
CSS_STYLES = """
<style>
.main { background-color: #f4f9f9; color: #000000; }
.sidebar .sidebar-content { background-color: #d1e7dd; }
.stTextInput textarea { color: #000000 !important; }
.stSelectbox div[data-baseweb="select"],
.stSelectbox option,
.stSelectbox div[role="listbox"] div {
color: black !important;
background-color: #d1e7dd !important;
}
.stSelectbox svg { fill: black !important; }
.main-title {
font-size: 88px;
font-weight: bold;
color: rgb(33, 238, 238);
}
.sub-title {
font-size: 100px;
color: #6B6B6B;
margin-top: -1px;
}
.stButton>button {
background-color: rgb(33, 225, 250);
color: white;
font-size: 69px;
}
.stImage img {
border-radius: 10px;
box-shadow: 2px 2px 10px rgba(0,0,0,0.1);
}
.logo {
text-align: center;
margin-bottom: 20px;
}
.report-container {
background-color: #ffffff;
border-radius: 15px;
padding: 25px;
margin-top: 20px;
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
border-left: 5px solid #21eeef;
}
.report-text {
font-family: 'Courier New', monospace;
font-size: 16px;
line-height: 1.6;
color: #2c3e50;
}
.download-btn {
background-color: #21eeef !important;
color: white !important;
border: none !important;
border-radius: 8px !important;
padding: 12px 24px !important;
}
</style>
"""
# ======================
# CORE FUNCTIONS
# ======================
def configure_application():
"""Initialize application settings and styling"""
st.set_page_config(**PAGE_CONFIG)
st.markdown(CSS_STYLES, unsafe_allow_html=True)
def initialize_api_client():
"""Create and validate Groq API client"""
load_dotenv()
api_key = os.getenv("GROQ_API_KEY")
if not api_key:
st.error("API key not found. Please verify .env configuration.")
st.stop()
return Groq(api_key=api_key)
def encode_logo(image_path):
"""Encode logo image to base64"""
try:
with open(image_path, "rb") as img_file:
return base64.b64encode(img_file.read()).decode("utf-8")
except FileNotFoundError:
st.error("Logo image not found! Using placeholder.")
return ""
def process_image_data(uploaded_file):
"""Convert image to base64 encoded string"""
try:
image = Image.open(uploaded_file)
buffer = io.BytesIO()
image.save(buffer, format=image.format)
return base64.b64encode(buffer.getvalue()).decode('utf-8'), image.format
except Exception as e:
st.error(f"Image processing error: {str(e)}")
return None, None
def generate_pdf_report(report_text):
"""Generate PDF document from report text"""
buffer = io.BytesIO()
doc = SimpleDocTemplate(buffer, pagesize=letter)
styles = getSampleStyleSheet()
story = []
# Add title
title = Paragraph("<b>Radiology Report</b>", styles['Title'])
story.append(title)
story.append(Spacer(1, 12))
# Add report content
content = Paragraph(report_text.replace('\n', '<br/>'), styles['BodyText'])
story.append(content)
doc.build(story)
buffer.seek(0)
return buffer
def generate_radiology_report(uploaded_file, client):
"""Generate AI-powered radiology analysis"""
base64_image, img_format = process_image_data(uploaded_file)
if not base64_image:
return None
image_url = f"data:image/{img_format.lower()};base64,{base64_image}"
try:
response = client.chat.completions.create(
model="llama-3.2-11b-vision-preview",
messages=[{
"role": "user",
"content": [
{"type": "text", "text": (
"As an AI radiologist, provide a detailed structured report including: "
"1. Imaging modality identification\n2. Anatomical structures visualized\n"
"3. Abnormal findings description\n4. Differential diagnoses\n"
"5. Clinical correlation recommendations"
)},
{"type": "image_url", "image_url": {"url": image_url}},
]
}],
temperature=0.2,
max_tokens=400,
top_p=0.5
)
return response.choices[0].message.content
except Exception as e:
st.error(f"API communication error: {str(e)}")
return None
# ======================
# UI COMPONENTS
# ======================
def display_main_interface():
"""Render primary application interface"""
# Encode logo image
logo_b64 = encode_logo("src/radiology.png")
# Center the logo and title using HTML and CSS
st.markdown(
f"""
<div style="text-align: center;">
<div class="logo">
<img src="data:image/png;base64,{logo_b64}" width="100">
</div>
<p class="main-title"> Radiology Analyzer</p>
<p class="sub-title">Advanced Medical Imaging Analysis</p>
</div>
""",
unsafe_allow_html=True
)
st.markdown("---")
# Action buttons
col1, col2 = st.columns([1, 1])
with col1:
if st.session_state.get('analysis_result'):
pdf_report = generate_pdf_report(st.session_state.analysis_result)
st.download_button(
label="πŸ“„ Download PDF Report",
data=pdf_report,
file_name="radiology_report.pdf",
mime="application/pdf",
use_container_width=True,
help="Download formal PDF version of the report",
key="download_pdf"
)
with col2:
if st.button("Clear Analysis πŸ—‘οΈ", use_container_width=True, help="Remove current results"):
st.session_state.pop('analysis_result')
st.rerun()
# Display analysis results
if st.session_state.get('analysis_result'):
st.markdown("### 🎯 Radiological Findings Report")
st.markdown(
f'<div class="report-container"><div class="report-text">{st.session_state.analysis_result}</div></div>',
unsafe_allow_html=True
)
def render_sidebar(client):
"""Create sidebar interface elements"""
with st.sidebar:
st.divider()
st.markdown("### Diagnostic Capabilities")
st.markdown("""
- **Multi-Modality Analysis**: X-ray, MRI, CT, Ultrasound
- **Pathology Detection**: Fractures, tumors, infections
- **Comparative Analysis**: Track disease progression
- **Structured Reporting**: Standardized output format
- **Clinical Correlation**: Suggested next steps
""")
st.divider()
st.subheader("Image Upload Section")
uploaded_file = st.file_uploader(
"Select Medical Image",
type=ALLOWED_FILE_TYPES,
help="Supported formats: PNG, JPG, JPEG"
)
if uploaded_file:
st.image(Image.open(uploaded_file),
caption="Uploaded Medical Image",
use_container_width=True)
if st.button("Initiate Analysis πŸ”", use_container_width=True):
with st.spinner("Analyzing image. This may take 20-30 seconds..."):
report = generate_radiology_report(uploaded_file, client)
st.session_state.analysis_result = report
st.rerun()
# ======================
# APPLICATION ENTRYPOINT
# ======================
def main():
"""Primary application controller"""
configure_application()
groq_client = initialize_api_client()
display_main_interface()
render_sidebar(groq_client)
if __name__ == "__main__":
main()