Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,250 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import requests
|
3 |
+
import json
|
4 |
+
from PIL import Image
|
5 |
+
import base64
|
6 |
+
from io import BytesIO
|
7 |
+
import pandas as pd
|
8 |
+
from datetime import datetime
|
9 |
+
import time
|
10 |
+
import logging
|
11 |
+
import os
|
12 |
+
from typing import Dict, Any, Optional
|
13 |
+
import re
|
14 |
+
from reportlab.lib import colors
|
15 |
+
from reportlab.lib.pagesizes import letter
|
16 |
+
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle
|
17 |
+
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
|
18 |
+
import io
|
19 |
+
import pytesseract # Tesseract OCR
|
20 |
+
from dotenv import load_dotenv # For .env file
|
21 |
+
|
22 |
+
# Load environment variables
|
23 |
+
load_dotenv()
|
24 |
+
|
25 |
+
# Configure logging
|
26 |
+
logging.basicConfig(
|
27 |
+
level=logging.INFO,
|
28 |
+
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
|
29 |
+
)
|
30 |
+
logger = logging.getLogger(__name__)
|
31 |
+
|
32 |
+
# Configuration and Constants
|
33 |
+
class Config:
|
34 |
+
GEMINI_URL = "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash-exp:generateContent"
|
35 |
+
GEMINI_API_KEY = os.getenv("GEMINI_API_KEY") # Load from .env
|
36 |
+
MAX_RETRIES = 3
|
37 |
+
TIMEOUT = 30
|
38 |
+
MAX_IMAGE_SIZE = (1600, 1600)
|
39 |
+
ALLOWED_MIME_TYPES = ["image/jpeg", "image/png"]
|
40 |
+
MAX_FILE_SIZE = 5 * 1024 * 1024 # 5MB
|
41 |
+
|
42 |
+
# Custom Exceptions
|
43 |
+
class APIError(Exception):
|
44 |
+
pass
|
45 |
+
|
46 |
+
class ImageProcessingError(Exception):
|
47 |
+
pass
|
48 |
+
|
49 |
+
# Initialize session state
|
50 |
+
def init_session_state():
|
51 |
+
if 'processing_history' not in st.session_state:
|
52 |
+
st.session_state.processing_history = []
|
53 |
+
if 'current_document' not in st.session_state:
|
54 |
+
st.session_state.current_document = None
|
55 |
+
if 'pdf_history' not in st.session_state:
|
56 |
+
st.session_state.pdf_history = []
|
57 |
+
|
58 |
+
# Page setup and styling
|
59 |
+
def setup_page():
|
60 |
+
st.set_page_config(
|
61 |
+
page_title="Medical Document Processor",
|
62 |
+
page_icon="🏥",
|
63 |
+
layout="wide",
|
64 |
+
initial_sidebar_state="expanded"
|
65 |
+
)
|
66 |
+
|
67 |
+
st.markdown("""
|
68 |
+
<style>
|
69 |
+
.main {padding: 2rem; max-width: 1200px; margin: 0 auto;}
|
70 |
+
.stCard {
|
71 |
+
background-color: white;
|
72 |
+
padding: 2rem;
|
73 |
+
border-radius: 10px;
|
74 |
+
box-shadow: 0 4px 6px rgba(0,0,0,0.1);
|
75 |
+
margin: 1rem 0;
|
76 |
+
}
|
77 |
+
.header-container {
|
78 |
+
background-color: #f8f9fa;
|
79 |
+
padding: 2rem;
|
80 |
+
border-radius: 10px;
|
81 |
+
margin-bottom: 2rem;
|
82 |
+
}
|
83 |
+
.stButton>button {
|
84 |
+
background-color: #007bff;
|
85 |
+
color: white;
|
86 |
+
border: none;
|
87 |
+
padding: 0.5rem 1rem;
|
88 |
+
border-radius: 5px;
|
89 |
+
transition: all 0.3s ease;
|
90 |
+
}
|
91 |
+
.stButton>button:hover {
|
92 |
+
background-color: #0056b3;
|
93 |
+
transform: translateY(-2px);
|
94 |
+
}
|
95 |
+
.element-container {opacity: 1 !important;}
|
96 |
+
.pdf-history-item {
|
97 |
+
background-color: #f8f9fa;
|
98 |
+
padding: 1rem;
|
99 |
+
border-radius: 8px;
|
100 |
+
margin: 0.5rem 0;
|
101 |
+
border: 1px solid #dee2e6;
|
102 |
+
}
|
103 |
+
.metric-card {
|
104 |
+
background-color: #f8f9fa;
|
105 |
+
padding: 1rem;
|
106 |
+
border-radius: 8px;
|
107 |
+
border: 1px solid #dee2e6;
|
108 |
+
margin: 0.5rem 0;
|
109 |
+
}
|
110 |
+
</style>
|
111 |
+
""", unsafe_allow_html=True)
|
112 |
+
|
113 |
+
# Rest of the code remains the same as in the previous version...
|
114 |
+
# (Include all classes and functions from the previous code here)
|
115 |
+
|
116 |
+
def main():
|
117 |
+
init_session_state()
|
118 |
+
setup_page()
|
119 |
+
|
120 |
+
st.title("🏥 Advanced Medical Document Processor")
|
121 |
+
st.markdown("Upload medical documents for automated processing and analysis.")
|
122 |
+
|
123 |
+
# Sidebar
|
124 |
+
with st.sidebar:
|
125 |
+
st.header("📋 Processing History")
|
126 |
+
if st.session_state.pdf_history:
|
127 |
+
for idx, pdf_record in enumerate(st.session_state.pdf_history):
|
128 |
+
with st.expander(f"Document {idx + 1}: {pdf_record['timestamp']}"):
|
129 |
+
st.download_button(
|
130 |
+
"📄 Download PDF",
|
131 |
+
pdf_record['data'],
|
132 |
+
file_name=pdf_record['filename'],
|
133 |
+
mime="application/pdf",
|
134 |
+
key=f"sidebar_{pdf_record['timestamp']}"
|
135 |
+
)
|
136 |
+
else:
|
137 |
+
st.info("No documents processed yet")
|
138 |
+
|
139 |
+
# Main content
|
140 |
+
uploaded_file = st.file_uploader(
|
141 |
+
"Choose a medical document",
|
142 |
+
type=['png', 'jpg', 'jpeg'],
|
143 |
+
help="Upload a clear image of a medical document (max 5MB)"
|
144 |
+
)
|
145 |
+
|
146 |
+
if uploaded_file:
|
147 |
+
try:
|
148 |
+
# Validate image
|
149 |
+
is_valid, message = ImageProcessor.validate_image(uploaded_file)
|
150 |
+
if not is_valid:
|
151 |
+
st.error(message)
|
152 |
+
return
|
153 |
+
|
154 |
+
# Display image
|
155 |
+
image = Image.open(uploaded_file)
|
156 |
+
col1, col2 = st.columns([1, 2])
|
157 |
+
|
158 |
+
with col1:
|
159 |
+
st.image(image, caption="Uploaded Document", use_column_width=True)
|
160 |
+
|
161 |
+
# Process document
|
162 |
+
if st.button("🔍 Process Document"):
|
163 |
+
with st.spinner("Processing document..."):
|
164 |
+
processor = DocumentProcessor()
|
165 |
+
results = processor.process_document(image)
|
166 |
+
|
167 |
+
# Generate PDF
|
168 |
+
pdf_bytes = PDFGenerator.create_pdf(results['structured_data'])
|
169 |
+
timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
|
170 |
+
pdf_filename = f"medical_report_{timestamp}.pdf"
|
171 |
+
|
172 |
+
# Store in session state
|
173 |
+
st.session_state.current_document = {
|
174 |
+
'timestamp': timestamp,
|
175 |
+
'results': results
|
176 |
+
}
|
177 |
+
st.session_state.processing_history.append(
|
178 |
+
st.session_state.current_document
|
179 |
+
)
|
180 |
+
st.session_state.pdf_history.append({
|
181 |
+
'timestamp': timestamp,
|
182 |
+
'filename': pdf_filename,
|
183 |
+
'data': pdf_bytes
|
184 |
+
})
|
185 |
+
|
186 |
+
# Display results
|
187 |
+
with col2:
|
188 |
+
st.success("Document processed successfully!")
|
189 |
+
st.markdown(f"**Document Type:** {results['document_type']}")
|
190 |
+
|
191 |
+
with st.expander("View Extracted Text"):
|
192 |
+
st.text_area(
|
193 |
+
"Raw Text",
|
194 |
+
results['extracted_text'],
|
195 |
+
height=200
|
196 |
+
)
|
197 |
+
|
198 |
+
# Display EHR View
|
199 |
+
if results['structured_data']:
|
200 |
+
EHRViewer.display_ehr(results['structured_data'])
|
201 |
+
|
202 |
+
# Download options
|
203 |
+
st.markdown("### 📥 Download Options")
|
204 |
+
col1, col2 = st.columns(2)
|
205 |
+
|
206 |
+
with col1:
|
207 |
+
json_str = json.dumps(results['structured_data'], indent=2)
|
208 |
+
st.download_button(
|
209 |
+
"⬇️ Download JSON",
|
210 |
+
json_str,
|
211 |
+
file_name=f"medical_data_{timestamp}.json",
|
212 |
+
mime="application/json"
|
213 |
+
)
|
214 |
+
|
215 |
+
with col2:
|
216 |
+
st.download_button(
|
217 |
+
"📄 Download PDF Report",
|
218 |
+
pdf_bytes,
|
219 |
+
file_name=pdf_filename,
|
220 |
+
mime="application/pdf"
|
221 |
+
)
|
222 |
+
|
223 |
+
# Display PDF History
|
224 |
+
st.markdown("### 📚 PDF History")
|
225 |
+
if st.session_state.pdf_history:
|
226 |
+
for pdf_record in st.session_state.pdf_history:
|
227 |
+
col1, col2 = st.columns([3, 1])
|
228 |
+
with col1:
|
229 |
+
st.write(f"Report from {pdf_record['timestamp']}")
|
230 |
+
with col2:
|
231 |
+
st.download_button(
|
232 |
+
"📄 View PDF",
|
233 |
+
pdf_record['data'],
|
234 |
+
file_name=pdf_record['filename'],
|
235 |
+
mime="application/pdf",
|
236 |
+
key=f"history_{pdf_record['timestamp']}"
|
237 |
+
)
|
238 |
+
else:
|
239 |
+
st.info("No PDF history available")
|
240 |
+
|
241 |
+
except Exception as e:
|
242 |
+
st.error(f"An error occurred: {str(e)}")
|
243 |
+
logger.exception("Error in main processing loop")
|
244 |
+
|
245 |
+
if __name__ == "__main__":
|
246 |
+
try:
|
247 |
+
main()
|
248 |
+
except Exception as e:
|
249 |
+
st.error("An unexpected error occurred. Please try again later.")
|
250 |
+
logger.exception("Unhandled exception in main application")
|