Merge pull request #399 from GSTT-CSC/378-geom-accuracy-wrong-slices

This update corrects the use of a mask created for slices 1 and 5, and taking the phantom centre from these slices rather than 7 in order to more accurately calculate geometric accuracy.

Below functions updated in the ACRobject.py are updated to more flexibly take the slice as input rather implicitly use slice 7:
- find_phantom_centre
- get_mask_image
- measure_orthogonal_lengths

Furthermore, for the get_mask_image function the threshold was increased to overcome background noise in some images.

hazenlib/ACRObject.py

@@ -16,10 +16,10 @@ def __init__(self, dcm_list):
         self.orientation_checks()
         # Determine whether image rotation is necessary
         self.rot_angle = self.determine_rotation()
+        # Find the centre coordinates of the phantom (circle) on slice 7 only:
+        self.centre, self.radius = self.find_phantom_center(self.images[6])
         # Store the DCM object of slice 7 as it is used often
         self.slice7_dcm = self.dcms[6]
-        # Find the centre coordinates of the phantom (circle)
-        self.centre, self.radius = self.find_phantom_center()
         # Store a mask image of slice 7 for reusability
         self.mask_image = self.get_mask_image(self.images[6])
 
@@ -144,17 +144,15 @@ def rotate_images(self):
             self.images, self.rot_angle, resize=False, preserve_range=True
         )
 
-    def find_phantom_center(self):
+    def find_phantom_center(self, img):
         """
-        Find the center of the ACR phantom by filtering the uniformity slice and using the Hough circle detector.
-
+        Find the center of the ACR phantom by filtering the input slice and using the Hough circle detector.
+        Args:
+            img (np.array): pixel array of the dicom
 
-        Returns
-        -------
-        centre  : tuple
-            Tuple of ints representing the (x, y) center of the image.
+        Returns:
+            tuple: Tuple of ints representing the (x, y) center of the image.
         """
-        img = self.images[6]
         dx, dy = self.pixel_spacing
 
         img_blur = cv2.GaussianBlur(img, (1, 1), 0)
@@ -174,14 +172,14 @@ def find_phantom_center(self):
         radius = int(detected_circles[2])
         return centre, radius
 
-    def get_mask_image(self, image, mag_threshold=0.05, open_threshold=500):
+    def get_mask_image(self, image, mag_threshold=0.07, open_threshold=500):
         """Create a masked pixel array
         Mask an image by magnitude threshold before applying morphological opening to remove small unconnected
         features. The convex hull is calculated in order to accommodate for potential air bubbles.
 
         Args:
-            image (_type_): _description_
-            mag_threshold (float, optional): magnitude threshold. Defaults to 0.05.
+            image (np.array): pixel array of the dicom
+            mag_threshold (float, optional): magnitude threshold. Defaults to 0.07.
             open_threshold (int, optional): open threshold. Defaults to 500.
 
         Returns:
@@ -240,7 +238,7 @@ def circular_mask(centre, radius, dims):
 
         return mask
 
-    def measure_orthogonal_lengths(self, mask):
+    def measure_orthogonal_lengths(self, mask, slice_index):
         """
         Compute the horizontal and vertical lengths of a mask, based on the centroid.
 
@@ -264,17 +262,18 @@ def measure_orthogonal_lengths(self, mask):
         """
         dims = mask.shape
         dx, dy = self.pixel_spacing
+        [(vertical, horizontal), radius] = self.find_phantom_center(self.images[slice_index])
 
-        horizontal_start = (self.centre[1], 0)
-        horizontal_end = (self.centre[1], dims[0] - 1)
+        horizontal_start = (horizontal, 0)
+        horizontal_end = (horizontal, dims[0] - 1)
         horizontal_line_profile = skimage.measure.profile_line(
             mask, horizontal_start, horizontal_end
         )
         horizontal_extent = np.nonzero(horizontal_line_profile)[0]
         horizontal_distance = (horizontal_extent[-1] - horizontal_extent[0]) * dx
 
-        vertical_start = (0, self.centre[0])
-        vertical_end = (dims[1] - 1, self.centre[0])
+        vertical_start = (0, vertical)
+        vertical_end = (dims[1] - 1, vertical)
         vertical_line_profile = skimage.measure.profile_line(
             mask, vertical_start, vertical_end
         )