Package sc.fiji.snt.analysis

Class PeripathDetector

java.lang.Object

sc.fiji.snt.analysis.PeripathDetector

public class PeripathDetector
extends Object

Detects intensity maxima (varicosities, spines, synaptic puncta) in annular cross-sections around traced paths. At each path node, a perpendicular cross-section is sampled from the image (see PathFitter), masked to an annulus defined by the node's radius (inner) and a configurable outer radius, and then analyzed for local maxima with prominence filtering via MaximumFinder. Detections from adjacent cross-sections are deduplicated via greedy non-maximum suppression.

Usage:

   PeripathDetector.Config cfg = new PeripathDetector.Config()
       .innerRadiusMultiplier(0.5)  // start at 0.5× node radius (closer to skeleton)
       .outerRadiusMultiplier(2.0)  // search up to 2× node radius
       .prominence(50)              // MaximumFinder noise tolerance
       .mergingDistance(1.5);        // NMS radius in calibrated units
   List<PeripathDetector.Detection> hits =
       PeripathDetector.detect(paths, image, cfg);
 

Author:

Tiago Ferreira

See Also:

• CrossSectionUtils
• PointDeduplicator

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final class	PeripathDetector.Config	Immutable configuration with builder-like setters.
static final class	PeripathDetector.Detection	One detected maximum (varicosity, spine, punctum, etc.).

Method Summary

Modifier and Type	Method	Description
static void	createTorusMask(Collection<Path> paths, ij.ImagePlus output, PeripathDetector.Config cfg, double fillValue)	Creates a binary torus mask around the given paths, writing into an ImagePlus.
static <T extends net.imglib2.type.numeric.RealType<T>> void	createTorusMask(Collection<Path> paths, net.imglib2.RandomAccessibleInterval<T> output, PeripathDetector.Config cfg, double fillValue)	Creates a binary torus mask around the given paths.
static <T extends net.imglib2.type.numeric.RealType<T>> List<PeripathDetector.Detection>	detect(Collection<Path> paths, net.imglib2.RandomAccessibleInterval<? extends net.imglib2.type.numeric.RealType<?>> img, PeripathDetector.Config cfg)	Detects intensity maxima along the given paths.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

detect

public static <T extends net.imglib2.type.numeric.RealType<T>>
List<PeripathDetector.Detection> detect(Collection<Path> paths,
 net.imglib2.RandomAccessibleInterval<? extends net.imglib2.type.numeric.RealType<?>> img,
 PeripathDetector.Config cfg)

Detects intensity maxima along the given paths.

Parameters:

paths - the paths to analyze (must have calibrated spacing)

img - the detection image (single channel). Coordinates are in the same space as the paths.

cfg - detection parameters

Returns:

list of detections, deduplicated and optionally assigned to nearest path

Throws:

IllegalArgumentException - if paths or image is null/empty

createTorusMask

public static <T extends net.imglib2.type.numeric.RealType<T>>
void createTorusMask(Collection<Path> paths,
 net.imglib2.RandomAccessibleInterval<T> output,
 PeripathDetector.Config cfg,
 double fillValue)

Creates a binary torus mask around the given paths. At each node, the annular search region (as defined by the PeripathDetector.Config) is back-projected into the output image, painting voxels inside the torus with the specified fill value.

Usage:

   // Create an empty mask matching the source image dimensions
   Img<UnsignedByteType> mask = ArrayImgs.unsignedBytes(width, height, depth);
   PeripathDetector.Config cfg = new PeripathDetector.Config()
       .innerRadiusMultiplier(0.8)
       .outerRadiusMultiplier(2.0);
   PeripathDetector.createTorusMask(paths, mask, cfg, 255);
 

Type Parameters:

T - pixel type of the output image

Parameters:

paths - the paths to generate the torus for

output - the output image to paint into (must be pre-allocated with dimensions matching the source image in pixel space); existing non-zero voxels are preserved

cfg - detection/annulus parameters (inner/outer radius settings)

fillValue - the value to write into torus voxels

Throws:

IllegalArgumentException - if paths or output is null/empty

createTorusMask

public static void createTorusMask(Collection<Path> paths,
 ij.ImagePlus output,
 PeripathDetector.Config cfg,
 double fillValue)

Creates a binary torus mask around the given paths, writing into an ImagePlus. This is a convenience overload of createTorusMask(Collection, RandomAccessibleInterval, Config, double) that accepts an IJ1 image directly.

Parameters:

paths - the paths to generate the torus for

output - the output image to paint into (must be pre-allocated with dimensions matching the source image in pixel space)

cfg - detection/annulus parameters (inner/outer radius settings)

fillValue - the value to write into torus voxels

Throws:

IllegalArgumentException - if paths or output is null/empty

See Also:

• createTorusMask(Collection, RandomAccessibleInterval, Config, double)