ToskaSkeleton class for skeleton analysis

This notebook demonstrates the usage of the ToskaSkeleton class for skeleton analysis with napari-toska

import napari_toska as nts
from skimage.morphology import skeletonize
from skimage.data import binary_blobs
from skimage.measure import label
import napari

viewer = napari.Viewer()

labels = label(binary_blobs(rng=0))
skeleton = skeletonize(labels)

Skeleton = nts.ToskaSkeleton(labels, neighborhood='n8')
Skeleton.analyze()

Building Graph: 100%|██████████| 48/48 [00:00<00:00, 1998.44it/s]

detected malformatted label: 87 changed type from 2 (branch) -> 1 (end point)
Found isolated nodes:  [24]
Could not connect isolated node:  24

Finding spines: 100%|██████████| 14/14 [00:00<00:00, 1555.34it/s]
Calculating summary features: 100%|██████████| 14/14 [00:00<00:00, 875.01it/s]

viewer.add_layer(Skeleton)

<ToskaSkeleton layer 'labels_data' at 0x2d69ba6dd90>

The measured features can be retrieved by inspecting the Skeleton.features attribute. The features are stored in a pandas DataFrame. Note that some measurements describe individual parts of the skeleton (such as the branch_length) whereas others summarize the entire Skeleton (e.g., n_cycle_basis).

Skeleton.features

	label	object_type	skeleton_id	branch_length	spine	n_branches	n_endpoints	n_branch_points	n_cycle_basis	n_possible_undirected_cycles
0	1	1	2	0.0	0	25	11	13	2.0	3.0
1	2	1	4	0.0	0	1	2	0	0.0	0.0
2	3	1	4	0.0	0	1	2	0	0.0	0.0
3	4	1	2	0.0	0	25	11	13	2.0	3.0
4	5	1	1	0.0	0	1	2	0	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...
102	103	3	2	0.0	0	25	11	13	2.0	3.0
103	104	3	2	0.0	0	25	11	13	2.0	3.0
104	105	3	2	0.0	0	25	11	13	2.0	3.0
105	106	3	12	0.0	0	3	3	1	0.0	0.0
106	107	3	2	0.0	0	25	11	13	2.0	3.0

107 rows × 10 columns

To retrieve summary statistics about the skeletons in a single dataframe, you can group the dataframe by the skeleton_id and aggregate all other values. In this example, all branch_lengths are summed up whereas the first entry is kept for the summary fields, as they are identical for every graph element, anyway:

# Group Skeletons by skeleton_id
grouped = Skeleton.features.groupby('skeleton_id').agg({'branch_length': 'sum', 'n_branches': 'first', 'n_endpoints': 'first', 'n_branch_points': 'first', 'skeleton_id': 'first'})

# display in single dataframe
grouped

	branch_length	n_branches	n_endpoints	n_branch_points	skeleton_id
skeleton_id
1	7.071068	1	2	0	1
2	865.498700	25	11	13	2
3	4.242641	1	2	0	3
4	11.313708	1	2	0	4
5	35.355339	1	2	0	5
6	22.627417	1	2	0	6
7	185.261977	9	6	4	7
8	28.284271	1	2	0	8
9	2.828427	1	2	0	9
10	0.000000	0	0	0	10
11	9.899495	1	2	0	11
12	118.793939	3	3	1	12
13	1.414214	1	2	0	13
14	0.000000	0	0	0	14
15	24.041631	1	2	0	15