add axis label to eval perfusion and compute correlation coeffiecient

mu_map/polar_map/eval_perfusion.py

@@ -165,6 +165,7 @@ def plot_corellation(
 
 if __name__ == "__main__":
     import argparse
+    from scipy.stats import pearsonr
 
     parser = argparse.ArgumentParser(
         description="Create visualizations to compare the perfusion of different polar maps",
@@ -208,7 +209,6 @@ if __name__ == "__main__":
     perfusion_dlac = perfusion_dlac[keys_segments].values
     perfusion_noac = perfusion_noac[keys_segments].values
 
-    print(f"APE:")
     fig, axs = plt.subplots(1, 2, figsize=(10, 4.5))
     for i, (perfusion, title) in enumerate(
         [(perfusion_dlac, "DLAC"), (perfusion_noac, "No AC")]
@@ -216,6 +216,9 @@ if __name__ == "__main__":
         if args.plot == "bland_altman":
             bland_altman(perfusion, perfusion_ctac, ax=axs[i])
             axs[i].set_title(title)
+            axs[i].set_xlabel("Mean Perfusion Score")
+            if i == 0:
+                axs[i].set_ylabel("Difference in Perfusion Score")
         else:
             v_min = min(
                 perfusion_ctac.min(), perfusion_dlac.min(), perfusion_noac.min()
@@ -232,7 +235,10 @@ if __name__ == "__main__":
             axs[i].set_ylabel("CTAC")
 
         _ape = absolute_percent_error(perfusion, perfusion_ctac)
-        print(f" - {title:>6}: {_ape.mean():.5f}±{_ape.std():.5f}")
+        r, p_value = pearsonr(perfusion.flatten(), perfusion_ctac.flatten())
+        print(f"{title:>6}")
+        print(f" -         APE: {_ape.mean():.5f}±{_ape.std():.5f}")
+        print(f" - Correlation: R={r:.5f}, (p={p_value:.5f})")
 
     same_limits(axs)
     plt.tight_layout()