update annotation notebook

napari-workshops/notebooks/manual_annotation.md

@@ -20,8 +20,7 @@ There are 3 main types of manual annotation that napari provides, each correspon
 - drawing 2D polygons to identify particular regions of interest with the **Shapes** layer
 - painting labels to provide a pixel-wise annotation of an image with the **Labels** layer
 
-This tutorial will explore these three manual annotations in **[napari](https://napari.org/)**, using that same data from the image visualization tutorial.
-
+This tutorial will explore these three manual annotations in napari, using the same data from the previous image visualization tutorial. Once again, we will try to highlight the bidirectional communication between the viewer and the Python kernel.
 
 ## Setup
 
@@ -45,23 +44,22 @@ from napari.utils import nbscreenshot
 viewer = napari.Viewer()
 ```
 
-In this notebook we will load our data directly into napari using out builtin plugin reader
+In this notebook we will use the same `cells3d` data set:
 
 ```{code-cell} ipython3
-:tags: [remove-output]
+from skimage.data import cells3d
 
-viewer.open('data/nuclei.tif')
-```
+image_data = cells3d()  # shape (60, 2, 256, 256)
 
-This loads the 3D data into the napari viewer. 
-Scrolling to the 30th z-slice should look as follows:
+membranes = image_data[:, 0, :, :]
+nuclei = image_data[:, 1, :, :]
 
-```{code-cell} ipython3
-:tags: [remove-cell]
-
-viewer.dims.current_step = (30, 0, 0)
+viewer.add_image(nuclei)
 ```
 
+This loads the 3D data into the napari viewer on the 30th slice. Let's take a screenshot
+for record keeping.
+
 ```{code-cell} ipython3
 nbscreenshot(viewer)
 ```
@@ -107,6 +105,7 @@ After annotation, my data looks as follows:
 ```{code-cell} ipython3
 :tags: [remove-cell]
 
+# this data is added to represent coordinates of points placed manually
 viewer.layers['dividing'].add([30, 88.1365794 , 54.39650486])
 viewer.layers['non-dividing'].add(
       [[30,  13.90840911,  27.02955319],
@@ -157,21 +156,29 @@ print('Number of non-diving cells:', len(viewer.layers['non-dividing'].data))
 viewer.layers['non-dividing'].data
 ```
 
-To save a `csv` file with these values for each layer you can use our builtin writer functionality. Note these csv files can easily be opened up into standard tools like pandas or excel for further analysis.
+To save a `csv` file with these values for each layer you can use our builtin writer functionality. Note these csv files can easily be opened up into standard tools like [`pandas`](https://pandas.pydata.org) or Excel for further analysis.
 
 ```{code-cell} ipython3
 # Save out Points layer data to a csv file
 #viewer.layers['dividing'].save('dividing.csv', plugin='builtins')
 #viewer.layers['non-dividing'].save('non-dividing.csv', plugin='builtins')
 ```
 
-Points layers also have a properties dictionary that would enable you to add other attributes like `volume` or `maximum-intensity` should you calculate those for each cell. You can learn more about the these advanced points annotations from the [tutorial](https://napari.org/tutorials/applications/annotate_points).
+Points layers also have a properties dictionary that would enable you to add other attributes like `volume` or `maximum-intensity` should you calculate those for each cell. You can learn more about the these advanced points annotations from the [tutorial](https://napari.org/stable/tutorials/applications/annotate_points).
 
 
 ## Drawing polygons around cells
 Another common task for research biologists and bioimage analysts is drawing polygons around regions of interest, in this case nuclei. These polygons might be used for segmentation and to quantify properties of interest.
 
-For this example we'll work with a 2D maximum intensity projection of our cells in order to keep things simple. We can take the data we've already loaded into napari and use it for the projection.
+```{note}
+At present (napari 0.4.19), napari does not support adding or editing shapes in 3D viewer mode. This 
+means they can only be drawn/edited on the orthogonal 2D slices.
+However, shapes can be added programmatically that have vertexes that are not on the 2D orthogonal planes
+and they will be properly rendered in 3D.
+```
+
+For the sake of this example, lets make a 2D maximum intensity projection of our cells in order to keep things simple. We can actually use the data we've already loaded into napari, because it's just a numpy array,
+and use it for the projection.
 
 ```{code-cell} ipython3
 # Take the maximum intensity projection of the cells
@@ -191,15 +198,11 @@ viewer.add_image(nuclei_mip);
 nbscreenshot(viewer)
 ```
 
-We can now add an empty new shapes layer from the GUI using the new shapes button (middle of the left panel, 2nd from the left with a polygon on it) or programatically from the notebook.
+We can now add an empty new shapes layer from the GUI using the new shapes button (middle of the left panel, 2nd from the left with a polygon on it) or programatically from the notebook. (Let's ensure we're in 2D mode.)
 
 ```{code-cell} ipython3
-:tags: [remove-cell]
-
 viewer.dims.ndisplay = 2
-```
 
-```{code-cell} ipython3
 viewer.add_shapes(name='nuclei outlines', face_color='red', edge_color='white', opacity=0.7)
 ```
 
@@ -214,6 +217,7 @@ We will draw some shapes with the polygon tool around a couple of different nucl
 ```{code-cell} ipython3
 :tags: [remove-cell]
 
+# this data is added to represent coordinates of shapes drawn manually
 p1 = [[142.12070325,  94.13481824],
       [150.36827772,  89.26125151],
       [163.86430869,  87.76169251],
@@ -276,7 +280,7 @@ These shapes, and the underlying image can be saved as an svg file using our ded
 #viewer.layers.save('nuclei-outlines.svg', plugin='napari-svg')
 ```
 
-Similarly to the points layer, we're working on adding support for properties dictionary to the shapes layer which would allow you to assign attributes to each shape and do things like adjust shape color based on them.
+Alternately, the shapes can be saved to a `csv` file just like we saved the points earlier.
 
 One common thing to use a shapes for is creating a binary mask or labels image where each pixel is assigned an integer label of the shape it is contained within, if any. napari provides some tooling to make these conversions easy.
 
@@ -304,7 +308,8 @@ nbscreenshot(viewer)
 
 ## Painting labels for pixel-wise annotations
 
-With the labels layer we can now make pixel-wise annotaions using a paintbrush, fill bucket, and eraser tools (see the row of buttons in the control panel in the top left of the viewer).
+With the labels layer we can now make pixel-wise annotaions using a paintbrush, fill bucket, 
+and eraser tools (see the row of buttons in the control panel in the top left of the viewer).
 
 Using these tools we can touch up any of the labels that we got from our polygon masks or draw entirely new ones.