Upload 3 files

app.py

@@ -307,7 +307,7 @@ def generate_fc_visualization(age, mpo, education, gender, handedness,
         except (NameError, AttributeError):
             # Fall back to old style visualization
             plt.figure(figsize=(10, 8))
-            plt.imshow(custom_fc_mat, cmap='coolwarm', vmin=-1, vmax=1)
+            plt.imshow(custom_fc_mat, cmap='RdBu_r', vmin=-1, vmax=1)
             plt.colorbar(label='Correlation')
             plt.title(f'FC Matrix: Age {age}, Gender {"M" if gender_val else "F"}, Aphasia Score: {predicted_aphasia_score:.1f}')
             plt.savefig(temp_img_path)
@@ -439,8 +439,8 @@ with gr.Blocks(title="Aphasia Prediction with FC Visualization") as demo:
     3. Optionally, override the model's prediction with your own custom score
     4. If the model is not trained, go to the "Train Model" tab to train it first
     
-    The heatmap shows correlations between brain regions. Yellow indicates positive correlations (regions that activate together),
-    green indicates neutral correlations, and blue indicates negative correlations (regions with opposing activation patterns).
+    The heatmap shows correlations between brain regions. Red indicates positive correlations (regions that activate together),
+    white indicates neutral correlations, and blue indicates negative correlations (regions with opposing activation patterns).
     """)
 
 if __name__ == "__main__":

osf_demovae_adapter.py

@@ -743,13 +743,13 @@ def plot_connectivity_matrix(fc_matrix, subject_id=None, save_path=None, show_la
         print("Matplotlib is required for plotting connectivity matrices")
         return None
     
-    # Create a custom colormap (yellow-green-blue)
-    colors = [(1, 1, 0),      # Yellow for positive correlations
-              (0, 1, 0),      # Green for moderate correlations  
+    # Create a custom colormap (red-white-blue)
+    colors = [(1, 0, 0),      # Red for positive correlations
+              (1, 1, 1),      # White for neutral correlations  
               (0, 0, 1)]      # Blue for negative correlations
     
     n_bins = 256  # Number of discrete colors
-    custom_cmap = LinearSegmentedColormap.from_list("YellowGreenBlue", colors, N=n_bins)
+    custom_cmap = LinearSegmentedColormap.from_list("RedWhiteBlue", colors, N=n_bins)
     
     # Create figure
     fig_size = 12 if show_labels else 8
@@ -789,14 +789,14 @@ def plot_connectivity_matrix(fc_matrix, subject_id=None, save_path=None, show_la
     cbar.set_label("Correlation Strength", rotation=270, labelpad=15)
     
     # Add annotations explaining the color scheme
-    fig.text(0.01, 0.01, "Color scheme: Yellow (positive correlation), Green (neutral), Blue (negative correlation)", 
+    fig.text(0.01, 0.01, "Color scheme: Red (positive correlation), White (neutral), Blue (negative correlation)", 
              fontsize=8, ha='left')
     
     # Add explanation of matrix content
     explanation = (
         "This matrix shows the functional connectivity between brain regions.\n"
         "Each cell represents the correlation of activity between two regions.\n"
-        "Positive values (yellow) indicate regions that activate together.\n"
+        "Positive values (red) indicate regions that activate together.\n"
         "Negative values (blue) indicate regions with opposite activation patterns."
     )