Metrics
This list is focused on default measurements available through SNT’s GUI and does not include specialized metrics available via its scripting API. Thus, it only captures a subset of the full repertoire of SNT metrics.
For each metric SNT retrieves the descriptive statistics Mix, Max, Mean, Standard Deviation (SD), Sum and N, which may lead to inevitable redundancy between measurements. E.g., when measuring Branch length for a particular cell, it is possible to retrieve the length of the smallest branch (Min), the longest (Max), the average and standard deviation of all branch lengths (Mean and SD), their total length (Sum), and number (N).
Metrics ported from published literature include their associated publication in the definition.
Branch contraction
A measure of straightness. The ratio between the Euclidean distance of a branch (i.e., Euclidean distance between the first and last node of the branch) and its path length. Range of values: ]0–1] (unitless). Lmeasure metric^{1}
See also: Path contraction
Branch fractal dimension
Also known has Hausdorff dimension. Defined as the slope obtained from the loglog plot of Path distance vs Euclidean distance, as implemented by Lmeasure following the definition of Marks & Burke (2007). It is only computed for branches defined by at least five nodes. Lmeasure metric^{1} described in: doi:10.1002/cne.21418
Branch length
The path length of a branch (i.e., the sum of all its internode distances)
See also: Path length
Branch mean radius
The average of the radii of the nodes defining a branch
Branch surface area
Estimated surface area^{2} of a branch computed from treating each internode segment as a conical frustum and summing the surface area of all frusta
Branch volume
Estimated volume^{2} of a branch computed from treating each internode segment as a conical frustum and summing the volume of all frusta
Cable length
The total path length of a structure, i.e., the sum of all internode distances of its paths
Complexity Indices
Complexity Indices are ratios of anatomical properties that summarize branching patterns. Typically, this type of descriptors have been created to summarize (early) neural development in vitro
Complexity index: ACI
Also known as “Axonal Complexity Index”. An index based on path orders, defined as \(\frac{\sum_{n=1}^{N} {Path\,order  1}}{N}\), with \(N\) being the total number of paths in the reconstruction. Described in: doi:10.1523/JNEUROSCI.443406.2007
Complexity index: DCI
Also known as “Dendritic Complexity Index”. An index based on the number of primary neurites, total arbor length, and the number and Strahlerorder of terminal branches. Described in: doi:10.1523/JNEUROSCI.192209928.1999
Convex hull
Defined as the polygon (2D) or the polyhedron (3D) enclosing a reconstruction
Convex hull: Boundary size
The perimeter of the 2D polygon or the surface area of the 3D polyhedron of the convex hull
Convex hull: Boxivity
The extent to which the convex hull approaches a rectangle (2D) or a cuboid (3D). Range of values: 0–1 (unitless)
Convex hull: Centroidroot distance
The distance between the root of a neuronal arbor and the centroid of its convex hull
Convex hull: Elongation
The caliper (also known as Feret) diameter of the convex hull
Convex hull: Roundness
The extent to which the convex hull approaches a circle (2D) or a sphere (3D). Range of values: 0–1 (unitless)
Convex hull: Size
Either the area of the 2D polygon, or the volume of the 3D polyhedron defining the convex hull
Depth
The depth of the bounding box embedding the structure being measured
Extension angle
The overall outgrowth direction of a branch,path, or neurite. It is obtained from the slope of a linear regression performed across all coordinates on either the XY, XZ, or ZY plane. Extension angles are expressed in degrees, under a [0360[ range
Height
The height of the bounding box embedding the structure being measured
HortonStrahler bifurcation ratio
The average bifurcation ratio of Strahler bifurcation ratios
HortonStrahler root number
The highest HortonStrahler number of a tree, i.e., the HortonStrahler number of its root node
Inner branches
Defined as the branches of highest Strahler order. Typically, these correspond to the most ‘internal’ branches of an arbor, in direct sequence from the root. Note that Primary branches are inner branches starting at the tree’s root
See also: Primary branches, Terminal branches
Inner branches: Extension angle XY
The extension angle of inner branches in the XY plane
Inner branches: Extension angle XZ
The extension angle of inner branches in the XZ plane
Inner branches: Extension angle ZY
The extension angle of inner branches in the ZY plane
Inner branches: Length
The length of inner branches
Internode angle
The angle (in degrees, 0360 range) between a node and its immediate neighbors. I.e., if node B is preceeded by node A and followed by node C, the internode angle at position B is defined as the angle between vecors AB and BC: \(\angle (\overrightarrow{AB}, \overrightarrow{BC})\)
Internode distance
The distance between nodes defining a branch or a Path. Can be retrieved as squared internode distance when faster computations are required
Longest shortest path
The longest graph geodesic. Considering a graphtheory tree, the longest shortest path corresponds to its graph diameter (also known as maximum geodesic, or longest graph geodesic). Can only be computed for structures that are valid mathematical trees.
Longest shortest path: Extension angle XY
Extension angle of longest shortest path in the XY plane
Longest shortest path: Extension angle XZ
Extension angle of longest shortest path in the XZ plane
Longest shortest path: Extension angle ZY
Extension angle of longest shortest path in the ZY plane
Longest shortest path: Length
Length of longest shortest path
Maximum geodesic
Same as Length of longest shortest path
No. of branch nodes (branch fragmentation)
The total number of nodes (and thus compartments) in a branch
No. of branch points
The total number (count) of branch points (also known as fork points)
No. of branches
The total number (count) of branches
No. of fitted paths
The total number (count) of fitted paths
No. of inner branches
The number of branches of highest Strahler order. Typically, these correspond to the most ‘internal’ branches of an arbor, in direct sequence from the root
No. of path nodes (path fragmentation)
The total number of nodes (and thus compartments) in a path
No. of paths
The total number (count) of paths defining a structure
No. of primary branches
The total number (count) of primary (or rootassociated) branches. Primary branches have origin in a tree’s root, extending to the closest branch point or endpoint, i.e., they are inner branches starting at the root. Note that a primary branch can also be terminal.
No. of spines/varicosities
Sum of all spine/varicosity markers in a structure
No. of spines/varicosities per path
Number of spines/varicosities associated with a path
No. of terminal branches
The total number (count) of branches ending at terminal endpoints (tips)
No. of tips
The total number (count) of terminal endpoints in a structure
No. of total nodes
The total number (count) of nodes in a structure
Node intensity values
The pixel intensity at each node location
Node radius
The radius at each node, typically obtained from fitting procedures
Partition asymmetry
Lmeasure metric^{1}. Computed at each bifurcation point of the structure being measured. Note that branch points with more than 2 children are ignored. Given \(n1, n2\) the number of tips on each side of a bifurcation point, Partition asymmetry is defined as: \(\frac{abs(n1n2)}{(n1+n22)}\).
Path channel
The color channel associated with a path (multidimensional images)
Path contraction
A measure of straightness of a path. See Branch contraction for definition
Path extension angle XY plane
Extension angle of a path in the XY plane
Path extension angle XZ plane
Extension angle of a path in the XZ plane
Path extension angle ZY plane
Extension angle of a path in the ZY plane
Path frame
The timepoint associated with a path (multidimensional images)
Path length
The sum of all internode distances in a path
See also: Branch length
Path mean radius
The average of the radii of the nodes defining a traced path
Path order
Path spine/varicosity density
The number (count) of spine/varicosity markers associated with a path, divided by its path length
Path surface area
Estimated surface area^{2} of a path computed from treating each internode segment as a conical frustum and summing the surface area of all frusta
Path volume
Estimated volume^{2} of a path computed from treating each internode segment as a conical frustum and summing the volume of all frusta
Persistence diagram
Persistence landscapes
Primary branches
Primary branches that have origin in a tree’s root, extending to the closest branch point or endpoint, i.e., inner branches starting at the root. Also known as rootassociated branches. Note that a primary branch can also be terminal
See also: Inner branches, Terminal branches
Primary branches: Extension angle XY plane
Extension angle of primary branches in the XY plane
Primary branches: Extension angle XZ plane
Extension angle of primary branches in the XZ plane
Primary branches: Extension angle ZY plane
Extension angle of primary branches in the ZY plane
Primary branches: Length
The length of primary branches
See also: Inner branches length, Terminal branches length
Remote bif. angles
The angle between each bifurcation point and its children in the simplified graph, which comprise either branch points or terminal nodes. Note that branch points with more than 2 children are ignored. Lmeasure metric^{1}
Sholl: Decay
The Sholl regression coefficient
Sholl: Degree of Polynomial fit
The polynomial degree used to fit the Sholl profile. See Sholl › Fitting functions
Sholl: Kurtosis
See Kurtosis in Sholl › Metrics based on sampled data
Sholl: Max
See Max inters. in Sholl › Metrics based on sampled data
Sholl: Max (fitted)
See Critical value in Sholl › Metrics based on fitted data
Sholl: Max (fitted) radius
See Critical radius in Sholl › Metrics based on fitted data
Sholl: Mean
See Mean inters. in Sholl › Metrics based on sampled data
Sholl: No. maxima
The number of times Max inters. occurs in a Sholl profile. See Max inters. in Sholl › Metrics based on sampled data
Sholl: No. secondary maxima
The number of times a secondary peak occurs in a Sholl profile. See Max inters. in Sholl › Metrics based on sampled data
Sholl: Ramification index
See Schoenen Ramification index in Sholl › Metrics based on sampled data
Sholl: Skewness
See Skewness in Sholl › Metrics based on sampled data
Sholl: Sum
See Sum inters. in Sholl › Metrics based on sampled data
Surface area
Treating each internode segment as a conical frustum, the sum of the surface areas^{2} of all frusta
Terminal branches
Branches ending at terminal endpoints (tips)
See also: Inner branches, Primary branches
Terminal branches: Extension angle XY plane
Extension angle of terminal branches in the XY plane
Terminal branches: Extension angle XZ plane
Extension angle of terminal branches in the XZ plane
Terminal branches: Extension angle ZY plane
Extension angle of terminal branches in the ZY plane
Terminal branches: Length
The sum of branch lengths of branches ending at terminal endpoints (tips)
See also: Inner branches length, Primary branches length
TMD
Topological Morphology Descriptor. See persistence homology analysis
Volume
Treating each internode segment as a conical frustum, the sum of the volume^{2} of all frusta
Width
The width of the bounding box embedding the structure being measured
X,Y,Z coordinates
Cartesian coordinates in the threedimensional space
Notes

This list does not include all of the specialized metrics provided by dedicated SNT modules, such as Strahler, Sholl, Persistence diagrams/landscapes, or Graphbased analysis

Some combinations of metrics/statistics may not be meaningful: e.g., when measuring a single cell, pairing cable length to SD will not be useful, since only one cable length value can be computed. In such cases, the Measurements table appends ‘[Single metric]’ to such data

Each of the 80+ metrics is represented by five statistical properties: minimum, maximum, mean, standard deviation and sum, resulting in a total of at least \(80\times 5\) features. Note that there is an intrinsic redundancy between these features: E.g., for a given cell, retrieving Branch length’s N is effectively the same as retrieving No. of branches

NaN values for a reported metric typically reflect undefined operations (e.g., division by zero), or the fact that the reconstruction being parsed is not a valid mathematical tree

Currently, volumerelated metrics do not take into account path fillings
Group Statistics
SNT assembles comparison reports and simple statistical reports (twosample ttest/oneway ANOVA) for up to six groups of cells. This is described in Comparing Reconstructions. Descriptive statistics of measurements can be obtained by running Summarize Existing Results in the Measurements dialog or by running Frequency/Distribution Analysis commands.
Glossary
Mesh
A polygon mesh defines the shape of a threedimensional polyhedral object. In neuronal anatomy, meshes define neuropil annotations, typically compartments of a reference brain atlas (e.g., the hippocampal formation in mammals, or mushroom bodies in insects)
Multidimensional image
An image with more than 3 dimensions (3D). Examples include fluorescent images associated with multiple fluorophores (multichannel) and images with a timedimension (timelapse videos). A 3D multichannel timelapse has 5 dimensions
Neurite
Same as neuronal process. Either an axon or a dendrite
Path
Can be defined as a sequence of branches, starting from soma or a branch point until a termination. In manual and assisted (semiautomated) tracing, neuronal arbors are traced using paths, not branches. Fitting algorithms that take into account voxel intensities can be used to refine the centerline coordinates of a path, typically to obtain more accurate curvatures. Fitting procedures can also be used to estimate the volume of the neurite(s) associated with a path
(Neuronal) morphometry
Quantification of neuronal morphology
Neuropil
Any area in the nervous system. The cellular tissue around neuronal processes
Outofcore
Software with outofcore capabilities is able to process data that is too large to fit into a computer’s main memory. SNT supports outofcore tracing via scripting.
Reconstruction
See Tracing
ROI
Region of Interest. Define specific parts of an image to be processed in image processing routines
Skeleton
A thinned version of a digitize shape (such as a neuronal reconstruction) or of a binary image
Tracing
A digital reconstruction of a neuron or neurite. The term predates computational neuroscience and reflects the manual ‘tracing’ on paper performed with camera lucida devices by early neuroanatomists
Volume rendering
A visualization technique for displaying image volumes (3D images) directly as 3D objects
This glossary was assembled from the supplementary note of SNT’s publication: doi:10.1038/s41592021011057