diff --git a/calcium_analysis/image_segmentation.py b/calcium_analysis/image_segmentation.py
index 479ab47..a2e33cf 100644
--- a/calcium_analysis/image_segmentation.py
+++ b/calcium_analysis/image_segmentation.py
@@ -3,14 +3,25 @@
 from scipy import ndimage as ndi
 from typing import Tuple
 
-def stack_to_roi(image_path: str, sd: float) -> Tuple[np.ndarray, np.ndarray, np.ndarray]:
+import numpy as np
+from skimage import filters, measure, segmentation, feature, io
+from scipy import ndimage as ndi
+from typing import Tuple
 
+def stack_to_roi(image_path: str, sd: float) -> Tuple[np.ndarray, np.ndarray, np.ndarray]:
     """
     Perform image segmentation on an image stack.
 
     This function loads a numpy array from a .npy file, performs 
     thresholding and edge detection, identifies connected components,
     and saves the results.
+
+    Parameters:
+    image_path (str): The path to the .npy file containing the image stack.
+    sd (float): The standard deviation used for image segmentation.
+
+    Returns:
+    Tuple[np.ndarray, np.ndarray, np.ndarray]: A tuple containing the labeled image, the mean image, and the binary image after morphological opening.
     """
 
     # Load the image stack
@@ -35,3 +46,26 @@ def stack_to_roi(image_path: str, sd: float) -> Tuple[np.ndarray, np.ndarray, np
 
     # Return the results
     return labels, I_mean, I_bw2
+
+    # Load the image stack
+    I_t = np.load(image_path)
+
+    # Compute the mean image
+    I_mean = np.mean(I_t, axis=2)
+
+    # Perform thresholding
+    I_bw1 = I_mean > filters.threshold_local(I_mean, block_size=15)
+
+    # Perform morphological opening
+    I_bw2 = ndi.binary_opening(I_bw1, structure=np.ones((3, 3)))
+
+    # Identify connected components
+    labels = measure.label(I_bw2)
+
+    # Save the results
+    np.save('I_mean.npy', I_mean)
+    np.save('I_bw2.npy', I_bw2)
+    np.save('labels.npy', labels)
+
+    # Return the results
+    return labels, I_mean, I_bw2