de.unihalle.informatik.MiToBo.apps.nuclei2D

Class NucleusSeparator2DPeakSearch

java.lang.Object

de.unihalle.informatik.Alida.operator.ALDOperator

de.unihalle.informatik.MiToBo.core.operator.MTBOperator

de.unihalle.informatik.MiToBo.apps.nuclei2D.NucleusSeparator2DAlgos

de.unihalle.informatik.MiToBo.apps.nuclei2D.NucleusSeparator2DPeakSearch

All Implemented Interfaces:

de.unihalle.informatik.Alida.datatypes.ALDConfigurationValidator, de.unihalle.informatik.Alida.operator.events.ALDOperatorExecutionProgressEventListener, EventListener

@ALDAOperator(genericExecutionMode=ALL,
              level=STANDARD,
              shortDescription="Operator for separating merged nuclei regions in images.")
 @ALDDerivedClass
public class NucleusSeparator2DPeakSearch
extends NucleusSeparator2DAlgos

Operator for separating merged nuclei regions in images.

This operator analyzes nuclei images and detects nuclei regions. The main focus is thereby on seperating nuclei regions which are merged by common segmentation techniques. Consequently, binary pre-segmentations of nuclei regions can be provided as input to the operator.
Anyway, if no binary image is coming along, the gray-scale input image will first be binarized by applying Otsu thresholding and some morphological post-processing steps. Indeed this a standard procedure which probably won't suit your needs - better use your own nucleus detector instead.

After binarization the resulting regions are further analyzed, i.e. the centers of present nuclei regions are determined (for details refer to Jochen's project thesis).

Given estimates for the nuclei regions snakes are initialized and run to detect the complete nuclei regions. As result extracted nuclei contours are provided as a set of polygons and overlayed to the greyscale image.

This operator has been written as part of Jochen's project in WS 2011/12.

Author:

Jochen Luechtrath, Birgit Moeller

Nested Class Summary

Nested classes/interfaces inherited from class de.unihalle.informatik.Alida.operator.ALDOperator

de.unihalle.informatik.Alida.operator.ALDOperator.HidingMode, de.unihalle.informatik.Alida.operator.ALDOperator.HistoryConstructionMode

Field Summary

Fields

Modifier and Type	Field and Description
private MTBImageByte	BWImg
private double	gamma Step-size in iterative snake optimization.
private Double	mad
private NucleusSeparator2DPeakSearch_RegionSeparator	nrs
private Float	peakdistance
private MTBImageRGB	resultRGBimage RGB image overlay with result nuclei contours.
private de.unihalle.informatik.Alida.datatypes.ALDDirectoryString	saveSnakeLoc Path where to save intermediate results.
private boolean	saveSnakes Flag to enable/disable saving of intermediate snake results.
private double	segmentlength Snake segment length.
private boolean	showSnakes Flag to enable/disable display of snake results.
private int	snakeIterations Maximal number of snake iterations in iterative optimization.
private MTBPolygon2DSet	snakes Set of resulting contours.
private Integer	suppressor

Fields inherited from class de.unihalle.informatik.MiToBo.apps.nuclei2D.NucleusSeparator2DAlgos

inputImg, labelImg, resultImg

Fields inherited from class de.unihalle.informatik.Alida.operator.ALDOperator

completeDAG, name, operatorExecutionEventlistenerList, portHashAccess, verbose, versionProvider

Constructor Summary

Constructors

Constructor and Description
NucleusSeparator2DPeakSearch() Default constructor.
NucleusSeparator2DPeakSearch(MTBImageShort grey) Constructor with input image.
NucleusSeparator2DPeakSearch(MTBImageShort grey, MTBImage label) Constructor with gray-scale and binary image.

Method Summary

Modifier and Type	Method and Description
void	configureSnakes(double g, int its, double avglength) Configures basic snake parameters.
String	getDocumentation()
MTBImageShort	getLabelImage() Returns gray-scale label image with nuclei regions.
MTBRegion2DSet	getNRSCandidates() Returns possible peak locations from NucleusRegionSeperator.
MTBImage	getNRSDistImg() Returns distance image calculated intermediately.
MTBPolygon2DSet	getNRSEllipsoidSnakePrimer() Creates and returns a set of polygons as starting regions for snake analysis based on hypothesized ellipses from nucleus information
MTBRegion2DSet[]	getNRSResultAreas() Returns set of result regions from region separator.
MTBRegion2DSet	getNRSResultCenters() Returns set of detected region centers from region separator.
MTBPolygon2DSet	getNRSVoronoidSnakePrimer() Creates and returns a set of polygons as starting regions for the snakes.
MTBRegion2DSet	getOtsuRegs() Returns set of regions resulting from Otsu thresholding.
MTBImageRGB	getResultImage() Returns RGB color image overlayed with nuclei contours.
MTBPolygon2DSet	getSnakes() Returns the result snakes.
protected void	operate()
void	setInImg(MTBImage inImg) Specify gray-scale input image.
void	setLabelImg(MTBImage label) Specify label input image.
void	setMad(double _mad) Sets maximal allowed discrepancy of direct connection to actual profile between two peaks.
void	setMax_d2c(float max_d2c) Sets factor to adjust maximal allowed distance from lower to upper peak.
void	setSavePath(String path) Sets the path to where intermediate snake results are to be written.
void	setSaveSnakes(boolean save) (De)activates saving of iteration step results.
void	setShowSnakes(boolean show) (De)activates displaying of intermediate results during snake-iterations.
void	setSnakeIterations(int its) Sets number of iterations for snake optimization.
void	setSuppressor(int sup) Sets threshold to suppress peak analysis close to scraggy contours.

Methods inherited from class de.unihalle.informatik.MiToBo.apps.nuclei2D.NucleusSeparator2DAlgos

getResultImg, setInputGrayScaleImage, setInputLabelImage, setResultImg

Methods inherited from class de.unihalle.informatik.MiToBo.core.operator.MTBOperator

readResolve

Methods inherited from class de.unihalle.informatik.Alida.operator.ALDOperator

addOperatorExecutionProgressEventListener, addParameter, addParameter, addParameterUnconditioned, fieldContained, fireOperatorExecutionProgressEvent, getALDPortHashAccessKey, getConstructionMode, getHidingMode, getInactiveParameterNames, getInInoutNames, getInInoutNames, getInNames, getInOutNames, getMissingRequiredInputs, getName, getNumParameters, getOutInoutNames, getOutNames, getParameter, getParameterDescriptor, getParameterDescriptorUnconditioned, getParameterNames, getParameterUnconditioned, getSupplementalNames, getVerbose, getVersion, handleOperatorExecutionProgressEvent, hasInOutParameters, hasParameter, isAnnotatedParameter, isConfigured, print, print, print, printInterface, printInterface, readHistory, reinitializeParameterDescriptors, removeOperatorExecutionProgressEventListener, removeParameter, runOp, runOp, runOp, setConstructionMode, setConstructionMode, setConstructionMode, setHidingMode, setName, setParameter, setParameterUnconditioned, setVerbose, toStringVerbose, unconfiguredItems, validate, validateCustom, validateGeneric, writeHistory, writeHistory, writeHistory

Methods inherited from class java.lang.Object

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

Field Detail

snakeIterations

@Parameter(label="Max. Snake Iterations",
           required=false,
           direction=IN,
           dataIOOrder=3,
           mode=STANDARD,
           description="Maximal number of snake iterations.")
private int snakeIterations

Maximal number of snake iterations in iterative optimization.

gamma

@Parameter(label="Snake Step Size",
           required=false,
           direction=IN,
           dataIOOrder=4,
           mode=STANDARD,
           description="Step size gamma for snake gradient-descent.")
private double gamma

Step-size in iterative snake optimization.

segmentlength

@Parameter(label="Snake Segment Length",
           required=false,
           direction=IN,
           dataIOOrder=5,
           mode=ADVANCED,
           description="Desired length of snake segments.")
private double segmentlength

Snake segment length.

showSnakes

@Parameter(label="Show Snakes",
           required=false,
           direction=IN,
           dataIOOrder=-10,
           mode=STANDARD,
           supplemental=true,
           description="Flag to enable/disable snake display.")
private boolean showSnakes

Flag to enable/disable display of snake results.

saveSnakes

@Parameter(label="Save Snakes",
           required=false,
           direction=IN,
           dataIOOrder=-9,
           mode=ADVANCED,
           supplemental=true,
           description="Flag to enable/disable saving of intermediate results.")
private boolean saveSnakes

Flag to enable/disable saving of intermediate snake results.

saveSnakeLoc

@Parameter(label="Save Results to...",
           required=false,
           direction=IN,
           dataIOOrder=-8,
           mode=ADVANCED,
           supplemental=true,
           description="Path to where intermediate results will be stored.")
private de.unihalle.informatik.Alida.datatypes.ALDDirectoryString saveSnakeLoc

Path where to save intermediate results.

snakes

@Parameter(label="Result Nuclei Contours",
           required=false,
           direction=OUT,
           description="Set of resulting snakes.")
private MTBPolygon2DSet snakes

Set of resulting contours.

resultRGBimage

@Parameter(label="Result Snake Image",
           required=false,
           direction=OUT,
           description="RGB image with nuclei contour overlay.")
private transient MTBImageRGB resultRGBimage

RGB image overlay with result nuclei contours.

mad

private transient Double mad

peakdistance

private transient Float peakdistance

suppressor

private transient Integer suppressor

BWImg

private transient MTBImageByte BWImg

nrs

private transient NucleusSeparator2DPeakSearch_RegionSeparator nrs

Constructor Detail

NucleusSeparator2DPeakSearch

public NucleusSeparator2DPeakSearch()
                             throws de.unihalle.informatik.Alida.exceptions.ALDOperatorException

Default constructor.

Throws:

de.unihalle.informatik.Alida.exceptions.ALDOperatorException

NucleusSeparator2DPeakSearch

public NucleusSeparator2DPeakSearch(MTBImageShort grey)
                             throws de.unihalle.informatik.Alida.exceptions.ALDOperatorException

Constructor with input image.

Parameters:

grey - Original greyscale nucleus image.

Throws:

de.unihalle.informatik.Alida.exceptions.ALDOperatorException

NucleusSeparator2DPeakSearch

public NucleusSeparator2DPeakSearch(MTBImageShort grey,
                                    MTBImage label)
                             throws de.unihalle.informatik.Alida.exceptions.ALDOperatorException

Constructor with gray-scale and binary image.

Parameters:

grey - Original greyscale nucleus image.

label - Label image of pre-segmented nuclei regions.

Throws:

de.unihalle.informatik.Alida.exceptions.ALDOperatorException

Method Detail

setInImg

public void setInImg(MTBImage inImg)

Specify gray-scale input image.

Parameters:

inImg - Gray-scale input image to be processed.

setLabelImg

public void setLabelImg(MTBImage label)

Specify label input image.

Parameters:

bw - Label input image.

setShowSnakes

public void setShowSnakes(boolean show)

(De)activates displaying of intermediate results during snake-iterations.

Parameters:

If - true, displaying results is enabled.

getSnakes

public MTBPolygon2DSet getSnakes()

Returns the result snakes.

Returns:

Resulting nuclei contours.

getResultImage

public MTBImageRGB getResultImage()

Returns RGB color image overlayed with nuclei contours.

Returns:

Color image with nuclei contours.

getLabelImage

public MTBImageShort getLabelImage()

Returns gray-scale label image with nuclei regions.

Returns:

Gray-scale image with nuclei regions.

setSaveSnakes

public void setSaveSnakes(boolean save)

(De)activates saving of iteration step results.

Parameters:

If - true, saving of results is enabled.

setSavePath

public void setSavePath(String path)

Sets the path to where intermediate snake results are to be written.

Parameters:

Path - in filesystem, folder must exist.

setSnakeIterations

public void setSnakeIterations(int its)

Sets number of iterations for snake optimization.

Parameters:

its - Number of iterations.

configureSnakes

public void configureSnakes(double g,
                            int its,
                            double avglength)

Configures basic snake parameters.

Parameters:

g - Array of snake step-sizes.

its - Number of iterations.

avglength - Desired length of snake segments.

getOtsuRegs

public MTBRegion2DSet getOtsuRegs()

Returns set of regions resulting from Otsu thresholding.

Returns:

Set of regions created by Otsu thresholding.

getNRSDistImg

public MTBImage getNRSDistImg()

Returns distance image calculated intermediately.

Returns:

Distance-transformation image as done by NucleusRegionSeparator.

getNRSVoronoidSnakePrimer

public MTBPolygon2DSet getNRSVoronoidSnakePrimer()
                                          throws de.unihalle.informatik.Alida.exceptions.ALDOperatorException,
                                                 de.unihalle.informatik.Alida.exceptions.ALDProcessingDAGException

Creates and returns a set of polygons as starting regions for the snakes.

The method is based on voronoi tesselation of Otsu-regions.

Returns:

Set of initial snake polygons.

Throws:

de.unihalle.informatik.Alida.exceptions.ALDProcessingDAGException

de.unihalle.informatik.Alida.exceptions.ALDOperatorException

getNRSEllipsoidSnakePrimer

public MTBPolygon2DSet getNRSEllipsoidSnakePrimer()

Creates and returns a set of polygons as starting regions for snake analysis based on hypothesized ellipses from nucleus information

getNRSResultAreas

public MTBRegion2DSet[] getNRSResultAreas()

Returns set of result regions from region separator.

Returns:

Splitted regions.

getNRSResultCenters

public MTBRegion2DSet getNRSResultCenters()

Returns set of detected region centers from region separator.

Returns:

Set of detected nuclei centers from NucleusRegionSeperator.

getNRSCandidates

public MTBRegion2DSet getNRSCandidates()

Returns possible peak locations from NucleusRegionSeperator.

Returns:

Set of possible peak locations from NucleusRegionSeperator.

setMax_d2c

public void setMax_d2c(float max_d2c)

Sets factor to adjust maximal allowed distance from lower to upper peak.

Parameters:

max_d2c - Maximal distance, default 1.5.

setMad

public void setMad(double _mad)

Sets maximal allowed discrepancy of direct connection to actual profile between two peaks.

Parameters:

mad - Maximal discrepancy, default 0.11.

setSuppressor

public void setSuppressor(int sup)

Sets threshold to suppress peak analysis close to scraggy contours.

Parameters:

sup - Threshold, should be > 4.

operate

protected void operate()
                throws de.unihalle.informatik.Alida.exceptions.ALDOperatorException,
                       de.unihalle.informatik.Alida.exceptions.ALDProcessingDAGException

Specified by:

operate in class NucleusSeparator2DAlgos

Throws:

de.unihalle.informatik.Alida.exceptions.ALDOperatorException

de.unihalle.informatik.Alida.exceptions.ALDProcessingDAGException

getDocumentation

public String getDocumentation()

Overrides:

getDocumentation in class de.unihalle.informatik.Alida.operator.ALDOperator