import gradio as gr
import os
import re
import logging
import tempfile
import shutil
import base64
from datetime import datetime
from PIL import Image
import html # Import the html module for escaping
from .patient_history import PatientHistoryManager, ReportGenerator
class UIComponents:
def __init__(self, auth_manager, database_manager, wound_analyzer):
self.auth_manager = auth_manager
self.database_manager = database_manager
self.wound_analyzer = wound_analyzer
self.current_user = {}
self.patient_history_manager = PatientHistoryManager(database_manager)
self.report_generator = ReportGenerator()
# Ensure uploads directory exists
if not os.path.exists("uploads"):
os.makedirs("uploads", exist_ok=True)
def image_to_base64(self, image_path):
"""Convert image to base64 data URL for embedding in HTML"""
if not image_path or not os.path.exists(image_path):
return None
try:
with open(image_path, "rb") as image_file:
encoded_string = base64.b64encode(image_file.read()).decode()
# Determine image format
image_ext = os.path.splitext(image_path)[1].lower()
if image_ext in [".jpg", ".jpeg"]:
mime_type = "image/jpeg"
elif image_ext == ".png":
mime_type = "image/png"
elif image_ext == ".gif":
mime_type = "image/gif"
else:
mime_type = "image/png" # Default to PNG
return f"data:{mime_type};base64,{encoded_string}"
except Exception as e:
logging.error(f"Error converting image to base64: {e}")
return None
def markdown_to_html(self, markdown_text):
"""Convert markdown text to proper HTML format with enhanced support"""
if not markdown_text:
return ""
# Escape HTML entities first to prevent issues with special characters
html_text = html.escape(markdown_text)
# Convert headers
html_text = re.sub(r"^### (.*?)$", r"\1 ", html_text, flags=re.MULTILINE)
html_text = re.sub(r"^## (.*?)$", r"\1 ", html_text, flags=re.MULTILINE)
html_text = re.sub(r"^# (.*?)$", r"\1 ", html_text, flags=re.MULTILINE)
# Convert bold text
html_text = re.sub(r"\*\*(.*?)\*\*", r"\1 ", html_text)
# Convert italic text
html_text = re.sub(r"\*(.*?)\*", r"\1 ", html_text)
# Convert code blocks (triple backticks)
html_text = re.sub(r"```(.*?)```", r"\1\1", html_text)
# Convert blockquotes
html_text = re.sub(r"^> (.*?)$", r"\1 ", html_text, flags=re.MULTILINE)
# Convert links
html_text = re.sub(r"\[(.*?)\]\((.*?)\)", r"\1 ", html_text)
# Convert horizontal rules
html_text = re.sub(r"^\s*[-*_]{3,}\s*$", r"")
in_list = True
result_lines.append(f"{stripped[2:]} ")
else:
if in_list:
result_lines.append(" ")
in_list = False
if stripped:
result_lines.append(f"{stripped}
")
else:
result_lines.append("
⚠️ IMPORTANT DISCLAIMER
This model is for testing and educational purposes only and is NOT a replacement for professional medical advice.
Information generated may be inaccurate. Always consult a qualified healthcare provider for medical concerns. This AI system uses chain-of-thought reasoning to show its decision-making process, but should never be used as the sole basis for clinical decisions.
Uploaded images may be stored and used for testing and model improvement purposes.
""")
# Main interface with conditional visibility (ORIGINAL STRUCTURE)
with gr.Row():
# Professional Authentication Panel (visible when not logged in)
with gr.Column(visible=True) as auth_panel:
gr.HTML("""
🏥 SmartHeal Access
Secure Healthcare Professional Portal
""")
with gr.Tabs():
with gr.Tab("🔐 Professional Login") as login_tab:
gr.HTML("""
Welcome Back
Access your professional dashboard
""")
login_username = gr.Textbox(
label="👤 Username",
placeholder="Enter your username"
)
login_password = gr.Textbox(
label="🔒 Password",
type="password",
placeholder="Enter your secure password"
)
login_btn = gr.Button(
"🚀 Sign In to Dashboard",
variant="primary",
size="lg"
)
login_status = gr.HTML(
value="Enter your credentials to access the system
"
)
with gr.Tab("📝 New Registration") as signup_tab:
gr.HTML("""
Create Account
Join the SmartHeal healthcare network
""")
signup_username = gr.Textbox(
label="👤 Username",
placeholder="Choose a unique username"
)
signup_email = gr.Textbox(
label="📧 Email Address",
placeholder="Enter your professional email"
)
signup_password = gr.Textbox(
label="🔒 Password",
type="password",
placeholder="Create a strong password"
)
signup_name = gr.Textbox(
label="👨⚕️ Full Name",
placeholder="Enter your full professional name"
)
signup_role = gr.Radio(
["practitioner", "organization"],
label="🏥 Account Type",
value="practitioner"
)
# Organization-specific fields
with gr.Group(visible=False) as org_fields:
gr.HTML("🏢 Organization Details ")
org_name = gr.Textbox(label="Organization Name", placeholder="Enter organization name")
phone = gr.Textbox(label="Phone Number", placeholder="Enter contact number")
country_code = gr.Textbox(label="Country Code", placeholder="e.g., +1, +44")
department = gr.Textbox(label="Department", placeholder="e.g., Emergency, Surgery")
location = gr.Textbox(label="Location", placeholder="City, State/Province, Country")
# Practitioner-specific fields
with gr.Group(visible=True) as prac_fields:
gr.HTML("🏥 Affiliation ")
organization_dropdown = gr.Dropdown(
choices=self.get_organizations_dropdown(),
label="Select Your Organization"
)
signup_btn = gr.Button(
"✨ Create Professional Account",
variant="primary",
size="lg"
)
signup_status = gr.HTML(
value="Fill in your details to create an account
"
)
# Practitioner Interface (hidden initially)
with gr.Column(visible=False) as practitioner_panel:
gr.HTML('👩⚕️ Practitioner Dashboard
')
user_info = gr.HTML("")
logout_btn_prac = gr.Button("🚪 Logout", variant="secondary", elem_classes=["logout-btn"])
# Main tabs for different functions
with gr.Tabs():
# WOUND ANALYSIS TAB
with gr.Tab("🔬 Wound Analysis"):
with gr.Row():
with gr.Column(scale=1):
gr.HTML("📋 Patient Information ")
patient_name = gr.Textbox(label="Patient Name", placeholder="Enter patient's full name")
patient_age = gr.Number(label="Age", value=30, minimum=0, maximum=120)
patient_gender = gr.Dropdown(
choices=["Male", "Female", "Other"],
label="Gender",
value="Male"
)
gr.HTML("🩹 Wound Information ")
wound_location = gr.Textbox(label="Wound Location", placeholder="e.g., Left ankle, Right arm")
wound_duration = gr.Textbox(label="Wound Duration", placeholder="e.g., 2 weeks, 1 month")
pain_level = gr.Slider(
minimum=0, maximum=10, value=5, step=1,
label="Pain Level (0-10)"
)
gr.HTML("⚕️ Clinical Assessment ")
moisture_level = gr.Dropdown(
choices=["Dry", "Moist", "Wet", "Saturated"],
label="Moisture Level",
value="Moist"
)
infection_signs = gr.Dropdown(
choices=["None", "Mild", "Moderate", "Severe"],
label="Signs of Infection",
value="None"
)
diabetic_status = gr.Dropdown(
choices=["Non-diabetic", "Type 1", "Type 2", "Gestational"],
label="Diabetic Status",
value="Non-diabetic"
)
with gr.Column(scale=1):
gr.HTML("📸 Wound Image Upload ")
wound_image = gr.Image(
label="Upload Wound Image",
type="filepath",
elem_classes=["image-upload"]
)
gr.HTML("📝 Medical History ")
previous_treatment = gr.Textbox(
label="Previous Treatment",
placeholder="Describe any previous treatments...",
lines=3
)
medical_history = gr.Textbox(
label="Medical History",
placeholder="Relevant medical conditions, surgeries, etc...",
lines=3
)
medications = gr.Textbox(
label="Current Medications",
placeholder="List current medications...",
lines=2
)
allergies = gr.Textbox(
label="Known Allergies",
placeholder="List any known allergies...",
lines=2
)
additional_notes = gr.Textbox(
label="Additional Notes",
placeholder="Any additional clinical observations...",
lines=3
)
analyze_btn = gr.Button("🔬 Analyze Wound", variant="primary", size="lg")
analysis_output = gr.HTML("")
# PATIENT HISTORY TAB
with gr.Tab("📋 Patient History"):
with gr.Row():
with gr.Column(scale=2):
gr.HTML("📊 Patient History Dashboard ")
history_btn = gr.Button("📋 Load Patient History", variant="primary")
patient_history_output = gr.HTML("")
with gr.Column(scale=1):
gr.HTML("🔍 Search Specific Patient ")
search_patient_name = gr.Textbox(
label="Patient Name",
placeholder="Enter patient name to search..."
)
search_patient_btn = gr.Button("🔍 Search Patient History", variant="secondary")
specific_patient_output = gr.HTML("")
# Interface already complete above - no additional tabs needed
# Event handlers
def handle_login(username, password):
user_data = self.auth_manager.authenticate_user(username, password)
if user_data:
self.current_user = user_data
return {
auth_panel: gr.update(visible=False),
practitioner_panel: gr.update(visible=True),
login_status: "✅ Login successful! Welcome to SmartHeal
"
}
else:
return {
login_status: "❌ Invalid credentials. Please try again.
"
}
def handle_signup(username, email, password, name, role, org_name, phone, country_code, department, location, organization_dropdown):
try:
if role == "organization":
org_data = {
'org_name': org_name,
'phone': phone,
'country_code': country_code,
'department': department,
'location': location
}
org_id = self.database_manager.create_organization(org_data)
user_data = {
'username': username,
'email': email,
'password': password,
'name': name,
'role': role,
'org_id': org_id
}
else:
# Extract org_id from dropdown selection
org_id = 1 # Default organization for now
user_data = {
'username': username,
'email': email,
'password': password,
'name': name,
'role': role,
'org_id': org_id
}
if self.auth_manager.create_user(user_data):
return {
signup_status: "✅ Account created successfully! Please login.
"
}
else:
return {
signup_status: "❌ Failed to create account. Username or email may already exist.
"
}
except Exception as e:
return {
signup_status: f"❌ Error: {str(e)}
"
}
def handle_analysis(patient_name, patient_age, patient_gender, wound_location, wound_duration,
pain_level, moisture_level, infection_signs, diabetic_status, previous_treatment,
medical_history, medications, allergies, additional_notes, wound_image):
try:
if not wound_image:
return "❌ Please upload a wound image for analysis.
"
# Show loading state
loading_html = """
Processing wound analysis...
"""
# 1. Construct questionnaire dictionary
questionnaire_data = {
'user_id': self.current_user.get('id'),
'patient_name': patient_name,
'patient_age': patient_age,
'patient_gender': patient_gender,
'wound_location': wound_location,
'wound_duration': wound_duration,
'pain_level': pain_level,
'moisture_level': moisture_level,
'infection_signs': infection_signs,
'diabetic_status': diabetic_status,
'previous_treatment': previous_treatment,
'medical_history': medical_history,
'medications': medications,
'allergies': allergies,
'additional_notes': additional_notes
}
# 2. Save questionnaire in DB
questionnaire_id = self.database_manager.save_questionnaire(questionnaire_data)
# 3. Run AI analysis with uploaded image
try:
# Log information about the wound image
if hasattr(wound_image, 'name'):
logging.info(f"Processing image: {wound_image.name}")
# First try direct analysis with the file-like object
analysis_result = self.wound_analyzer.analyze_wound(wound_image, questionnaire_data)
except Exception as e:
logging.error(f"AI analysis error (first attempt): {e}")
try:
# If that fails, try with PIL image
from PIL import Image
import io
# Reset file pointer if possible
if hasattr(wound_image, 'seek'):
wound_image.seek(0)
# Convert to PIL Image
pil_image = Image.open(wound_image)
analysis_result = self.wound_analyzer.analyze_wound(pil_image, questionnaire_data)
except Exception as e2:
logging.error(f"AI analysis error (second attempt): {e2}")
# Return error information for display
return f"""
❌ Analysis Error
There was an error analyzing the wound image:
{str(e)}\n{str(e2) if 'e2' in locals() else ''}
Please try again with a different image or contact support.
"""
# 4. Save AI analysis result
self.database_manager.save_analysis_result(questionnaire_id, analysis_result)
# 5. Save wound image metadata
if isinstance(wound_image, str):
image_url = wound_image
elif hasattr(wound_image, 'name'):
image_url = wound_image.name
else:
image_url = 'unknown'
image_data = {
'image_url': image_url,
'filename': os.path.basename(image_url),
'file_size': None,
'width': None,
'height': None
}
# 6. Format analysis results with visualization
formatted_analysis = self._format_analysis_results(analysis_result, image_url)
# 7. Generate HTML professional report for complete analysis
professional_report = self.report_generator.generate_analysis_report(
questionnaire_data,
analysis_result,
image_data.get('image_url')
)
return formatted_analysis + professional_report
except Exception as e:
logging.error(f"Analysis error: {e}")
return f"❌ Analysis failed: {str(e)}
"
def handle_logout():
self.current_user = {}
return {
auth_panel: gr.update(visible=True),
practitioner_panel: gr.update(visible=False)
}
def toggle_role_fields(role):
if role == "organization":
return {
org_fields: gr.update(visible=True),
prac_fields: gr.update(visible=False)
}
else:
return {
org_fields: gr.update(visible=False),
prac_fields: gr.update(visible=True)
}
def load_patient_history():
try:
user_id = self.current_user.get('id')
if not user_id:
return "❌ Please login first.
"
history_data = self.patient_history_manager.get_user_patient_history(user_id)
formatted_history = self.patient_history_manager.format_history_for_display(history_data)
return formatted_history
except Exception as e:
logging.error(f"Error loading patient history: {e}")
return f"❌ Error loading history: {str(e)}
"
def search_specific_patient(patient_name):
try:
user_id = self.current_user.get('id')
if not user_id:
return "❌ Please login first.
"
if not patient_name.strip():
return "⚠️ Please enter a patient name to search.
"
patient_data = self.patient_history_manager.search_patient_by_name(user_id, patient_name.strip())
if patient_data:
formatted_data = self.patient_history_manager.format_patient_data_for_display(patient_data)
return formatted_data
else:
return f"⚠️ No records found for patient: {patient_name}
"
except Exception as e:
logging.error(f"Error searching patient: {e}")
return f"❌ Error searching patient: {str(e)}
"
# Bind event handlers
login_btn.click(
handle_login,
inputs=[login_username, login_password],
outputs=[auth_panel, practitioner_panel, login_status]
)
signup_btn.click(
handle_signup,
inputs=[signup_username, signup_email, signup_password, signup_name, signup_role,
org_name, phone, country_code, department, location, organization_dropdown],
outputs=[signup_status]
)
signup_role.change(
toggle_role_fields,
inputs=[signup_role],
outputs=[org_fields, prac_fields]
)
analyze_btn.click(
handle_analysis,
inputs=[patient_name, patient_age, patient_gender, wound_location, wound_duration,
pain_level, moisture_level, infection_signs, diabetic_status, previous_treatment,
medical_history, medications, allergies, additional_notes, wound_image],
outputs=[analysis_output]
)
logout_btn_prac.click(
handle_logout,
outputs=[auth_panel, practitioner_panel]
)
history_btn.click(
load_patient_history,
outputs=[patient_history_output]
)
search_patient_btn.click(
search_specific_patient,
inputs=[search_patient_name],
outputs=[specific_patient_output]
)
return app
def _format_analysis_results(self, analysis_result, image_url=None):
"""Format analysis results for HTML display with base64 encoded images, always showing segmentation overlay."""
try:
# Extract key results
summary = analysis_result.get('summary', 'Analysis completed')
wound_detection = analysis_result.get('wound_detection', {})
segmentation_result = analysis_result.get('segmentation_result', {})
risk_assessment = analysis_result.get('risk_assessment', {})
recommendations = analysis_result.get('recommendations', '')
comprehensive_report = analysis_result.get('comprehensive_report', '')
# Detection metrics
detection_confidence = 0.0
wound_type = "Unknown"
length_cm = breadth_cm = area_cm2 = 0
if wound_detection.get('status') == 'success' and wound_detection.get('detections'):
detections = wound_detection.get('detections', [])
if detections:
detection_confidence = detections[0].get('detection_confidence', 0.0)
wound_type = detections[0].get('wound_type', 'Unknown')
length_cm = detections[0].get('length_cm', 0)
breadth_cm = detections[0].get('breadth_cm', 0)
area_cm2 = detections[0].get('surface_area_cm2', 0)
risk_level = risk_assessment.get('risk_level', 'Unknown')
risk_score = risk_assessment.get('risk_score', 0)
risk_factors = risk_assessment.get('risk_factors', [])
# Set risk class for styling
risk_class = "low"
if risk_level.lower() == "moderate":
risk_class = "moderate"
elif risk_level.lower() == "high":
risk_class = "high"
# Format risk factors
risk_factors_html = "" + "".join(f"{factor} " for factor in risk_factors) + " " if risk_factors else "No specific risk factors identified.
"
# Format guideline recommendations
guideline_recommendations = analysis_result.get('guideline_recommendations', [])
recommendations_html = "" + "".join(f"{rec} " for rec in guideline_recommendations if rec and len(rec) > 10) + " " if guideline_recommendations else "No specific recommendations available.
"
# --- Detection image ---
detection_image_base64 = None
if "overlay_path" in wound_detection and os.path.exists(wound_detection["overlay_path"]):
detection_image_base64 = self.image_to_base64(wound_detection["overlay_path"])
elif comprehensive_report:
detection_match = re.search(r"Detection Image: (.+?)(?:\n|$)", comprehensive_report)
if detection_match and detection_match.group(1) != "Not available" and os.path.exists(detection_match.group(1).strip()):
detection_image_base64 = self.image_to_base64(detection_match.group(1).strip())
detection_image_html = ""
if detection_image_base64:
detection_image_html = f"""
Wound Detection Visualization
"""
detection_html = f"""
🔍 Wound Detection & Classification
{detection_image_html}
Detection Confidence
{detection_confidence:.1%}
Total Wounds Detected
{wound_detection.get('total_wounds', 0)}
""" if wound_detection.get('status') == 'success' else f"""
Detection Status: FailedReason: {wound_detection.get('message', 'Unknown error')}
"""
# --- Segmentation overlay: prefer direct result! ---
segmentation_image_base64 = None
if "overlay_pil" in segmentation_result and isinstance(segmentation_result["overlay_pil"], Image.Image):
segmentation_image_base64 = pil_to_base64(segmentation_result["overlay_pil"])
elif "overlay_path" in segmentation_result and os.path.exists(segmentation_result["overlay_path"]):
segmentation_image_base64 = self.image_to_base64(segmentation_result["overlay_path"])
elif comprehensive_report:
segmentation_match = re.search(r"Segmentation Image: (.+?)(?:\n|$)", comprehensive_report)
if segmentation_match and segmentation_match.group(1) != "Not available" and os.path.exists(segmentation_match.group(1).strip()):
segmentation_image_base64 = self.image_to_base64(segmentation_match.group(1).strip())
segmentation_image_html = ""
if segmentation_image_base64:
segmentation_image_html = f"""
Wound Segmentation Visualization
"""
segmentation_html = f"""
📏 Wound Measurements
{segmentation_image_html}
Length
{length_cm:.2f} cm
Width
{breadth_cm:.2f} cm
Surface Area
{area_cm2:.2f} cm²
""" if segmentation_result.get('status') == 'success' else f"""
Segmentation Status: FailedReason: {segmentation_result.get('message', 'Unknown error')}
"""
# --- Main input image preview ---
image_visualization = ""
if image_url and os.path.exists(image_url):
image_base64 = self.image_to_base64(image_url)
if image_base64:
image_visualization = f"""
🖼️ Wound Image
Analysis completed successfully
"""
# --- Comprehensive report as HTML ---
comprehensive_report_html = ""
if comprehensive_report:
report_without_images = re.sub(r'## Analysis Images.*?(?=##|$)', '', comprehensive_report, flags=re.DOTALL)
comprehensive_report_html = self.markdown_to_html(report_without_images)
# --- Final Output ---
html_output = f"""
🔬 SmartHeal AI Analysis Results
Advanced Computer Vision & Medical AI Assessment
Analysis Summary: {summary}
{image_visualization}
{detection_html}
{segmentation_html}
⚠️ Risk Assessment
{risk_level} RISK
Risk Score: {risk_score}/10
Identified Risk Factors:
{risk_factors_html}
💡 Clinical Recommendations
{recommendations_html}
⚠️ Note: These recommendations are generated by AI and should be verified by healthcare professionals.
{f'
📋 Comprehensive Report {comprehensive_report_html}
' if comprehensive_report_html else ''}
Analysis completed by SmartHeal AI - Advanced Wound Care Assistant
"""
return html_output
except Exception as e:
logging.error(f"Error formatting results: {e}")
return f"❌ Error displaying results: {str(e)}
"