feat: Display uploaded X-ray image as reference before visualization

When users upload an X-ray image, immediately display it in the
visualization panel as a reference. Once segmentation or grounding
is complete, the visualization replaces the original image.

Flow:
1. User uploads X-ray → Shows original image on right
2. User requests segmentation → Original image visible as reference
3. Segmentation completes → Overlay replaces original image

This provides visual context while the agent processes the request.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -160,7 +160,7 @@ def get_or_create_agent(mode):
         )
     return agents_cache[mode]
 
-def chat(message, history, mode):
+def chat(message, history, mode, uploaded_image_path=None):
     """Chat function that uses the appropriate agent based on mode."""
     config = {"configurable": {"thread_id": f"thread_{mode}"}}
 
@@ -169,12 +169,14 @@ def chat(message, history, mode):
 
     # Handle multimodal input - Gemini 2.0 Flash supports vision
     image_content = None
+    current_upload = None  # Track the current uploaded image
     if isinstance(message, dict):
         text = message.get("text", "")
         files = message.get("files", [])
 
         if files and len(files) > 0:
             image_path = files[0]
+            current_upload = image_path  # Store for visualization
 
             # Check if it's a DICOM file
             is_dicom = image_path.lower().endswith(('.dcm', '.dicom'))
@@ -260,6 +262,10 @@ def chat(message, history, mode):
                     assistant_message = "I've segmented the requested anatomical structures. The visualization is shown on the right."
                 elif "grounding" in latest_viz:
                     assistant_message = "I've highlighted the requested regions. The visualization is shown on the right."
+    else:
+        # No visualization generated - show the uploaded image as reference
+        if current_upload:
+            viz_image = current_upload
 
     # Final fallback for empty messages
     if not assistant_message or assistant_message.strip() == "":