#include <ImFusion/US/BoneSegmentationAlgorithm.h>

Algorithm which segments bone in an ultrasound sweep. More...

Inheritance diagram for BoneSegmentationAlgorithm:

Detailed Description

Algorithm which segments bone in an ultrasound sweep.

This algorithm supports bone detection and segmentation. Many side output parameters can be acquired via the getters.

Classes
class	SweepDetectionResults
	Data structure to store the detections results for a given sweep. More...

Public Member Functions
	BoneSegmentationAlgorithm (const std::vector< UltrasoundSweep * > &sweeps)

void	configure (const Properties *p) override
	Configure this object instance by de-serializing the given Properties.

void	configuration (Properties *p) const override
	Serialize the current object configuration into the given Properties object.

void	computeCompoundedProbability ()
	Execute the detection on whole volume.

void	computeDetection (int slice)
	Execute only the detection.

void	computeSegmentation ()
	Execute only the segmentation (assumes that the detection has already been performed)

virtual void	detectOutliers ()
	Estimate how reliable the segmentations are.

void	compute () override
	Computes bone segmentation and prepares it for output.

OwningDataList	takeOutput () override
	Get the calculated results.

UltrasoundSweep *	sweep () const

int	bestSlice (UltrasoundSweep *sweep) const

bool	isResultAvailable (int slice, UltrasoundSweep *sweep) const

const std::vector< vec3 > &	detectedBone (int slice, UltrasoundSweep *sweep) const
	Returns the list of detected bone centerline points for a specific ultrasound slice.

const std::vector< vec3 > &	segmentedBone (int slice, UltrasoundSweep *sweep) const
	Returns the segmented bone centerline points for a specific ultrasound slice.

const std::vector< double > &	outlierness (UltrasoundSweep *sweep) const

const std::vector< double > &	detectionConfidence (UltrasoundSweep *sweep) const

const std::vector< std::vector< double > > &	boneClassification (UltrasoundSweep *sweep) const

std::vector< std::unique_ptr< PointCloud > >	extractPointClouds (int exportSkip, double outliernessMaxThreshold, std::vector< double > *outliernessOut) const

Public Member Functions inherited from BoneSegmentationBaseAlgorithm
	BoneSegmentationBaseAlgorithm (SharedImageSet images, SharedImageSet labels=nullptr)

void	configure (const Properties *p) override
	Configure this object instance by de-serializing the given Properties.

void	configuration (Properties *p) const override
	Serialize the current object configuration into the given Properties object.

void	compute () override
	Execute the algorithm.

OwningDataList	takeOutput () override
	Return any new Data that was created by the Algorithm during the last call to compute().

virtual bool	useDefaultAlgorithmController () const
	If you need to use a different controller in a derived class, set this to false.

virtual void	reset ()
	Clears current output.

const std::vector< vec3 > &	detectedBone () const

const std::vector< vec3 > &	segmentedBone () const

float	confidenceScore () const

PredictionMap	computePredictionMap (const SharedImage &curImage) const

DetectionResult	computeDetectedBone (const SharedImage &probabilityMap) const

std::vector< vec3 >	computeSegmentedBone (const SharedImage &curImage, const std::vector< vec3 > &detectedBone) const

Public Member Functions inherited from Algorithm
	Algorithm ()
	Default constructor will registers a single "compute" action that calls compute() and returns status().

virtual void	setProgress (Progress *progress)
	Sets a Progress interface the algorithm can use to notify observers about its computing progress.

Progress *	progress () const
	Returns the progress interface if set.

virtual int	status () const
	Indicates the status of the last call to compute().

virtual bool	survivesDataDeletion (const Data *) const
	Indicates whether the algorithm can handle (partial) deletion of the specified data, by default this checks whether the data is in the input list.

const FactoryInfo &	factoryInfo () const
	Returns the record describing how this Algorithm was instantiated by the AlgorithmFactory.

void	setFactoryInfo (const FactoryInfo &value)
	Sets the record describing how this Algorithm was instantiated by the AlgorithmFactory.

Status	runAction (const std::string &id)
	Run the action with name `id` if it exists.

const std::vector< Action > &	actions ()
	Get a mapping from Action id to Action as registered in this algorithm.

Public Member Functions inherited from Configurable
virtual void	configureDefaults ()
	Retrieve the properties of this object, replaces values with their defaults and sets it again.

void	registerParameter (ParameterBase *param)
	Register the given Parameter or SubProperty, so that it will be configured during configure()/configuration().

void	unregisterParameter (const ParameterBase *param)
	Remove the given Parameter or SubProperty from the list of registered parameters.

	Configurable (const Configurable &rhs)

	Configurable (Configurable &&rhs) noexcept

Configurable &	operator= (const Configurable &)

Configurable &	operator= (Configurable &&) noexcept

Static Public Member Functions
static bool	createCompatible (const DataList &data, Algorithm **a=0)

Static Public Member Functions inherited from BoneSegmentationBaseAlgorithm
static bool	createCompatible (const DataList &data, Algorithm **a=0)

Static Public Member Functions inherited from Algorithm
static bool	createCompatible (const DataList &data, Algorithm **a=nullptr)
	Factory function to check algorithm compatibility with input data and optionally instantiate it.

Protected Member Functions
void	init () override
	Clear/resize all vectors of images/results and initialize the algorithm (e.g. load the ML model)

bool	automaticDetection ()
	Find the best slice and compute its bone detection.

int	findBestSlice ()
	Compute the probability map on a subset of slices to find the one with the highest confidence.

std::shared_ptr< SharedImage >	boneProbability (int slice)
	Get the bone probability map in a given slice.

void	computeCompoundedProbabilityImpl (UltrasoundSweep *)

Protected Member Functions inherited from BoneSegmentationBaseAlgorithm
virtual void	updatePrediction ()
	Update prediction using current features.

virtual void	detection ()
	Run detection using ML prediction.

virtual void	legacyDetection ()
	Run legacy algorithm for bone detection.

virtual void	tracking ()
	Run tracking using last detection.

Protected Member Functions inherited from Algorithm
void	loadDefaults ()

void	registerAction (const std::string &id, const std::string &guiName, const std::function< Algorithm::Status(void)> &action)
	Register an action to be run via runAction.

template<typename D>
void	registerAction (const std::string &id, const std::string &guiName, Algorithm::Status(D::*action)(void))
	Template version of runAction that can be used with a pointer to a member function.

void	registerAction (const Action &action)
	Register an action.

Input
std::vector< UltrasoundSweep * >	m_sweeps
	Input ultrasound sweeps.

UltrasoundSweep *	m_sweep
	Currently analyzed sweep.

Output
std::map< UltrasoundSweep *, SweepDetectionResults >	m_results

std::vector< std::unique_ptr< PointCloud > >	m_outPointClouds

std::unique_ptr< SharedImage >	m_compoundedProba
	Compounded probability map for the bone.

Configurable Parameters
int	m_sliceStep
	Slice step for detection and propagation. Minimum value is 1.

bool	m_intermediateOutput
	Export intermediate images.

bool	m_exportPointCloud

int	m_pointCloudSkip
	The skip step for the exported point cloud.

bool	m_exportAllRelativePointClouds

bool	m_exportFrameBasedPointCloud

Outliers
Properties	m_outlierDetectionProperties

double	m_pointCloudMaxOutlierness

Additional Inherited Members
Public Types inherited from Algorithm
enum	Status { Unknown = -1 , Success = 0 , Error = 1 , InvalidInput , IncompleteInput , OutOfMemoryHost , OutOfMemoryGPU , UnsupportedGPU , UnknownAction , AbortedByUser , User = 1000 }
	Status codes. More...

Public Attributes inherited from Algorithm
Signal	signalOutputChanged
	Signal should be emitted by Algorithms when their output/result has changed.

Signal	signalParametersChanged
	Signal should be emitted by Algorithms when their parameter configuration has changed.

Public Attributes inherited from Configurable
Signal	signalParametersChanged
	Emitted whenever one of the registered Parameters' or SubPropertys' `signalValueChanged` signal was emitted.

Protected Attributes inherited from BoneSegmentationBaseAlgorithm
SharedImage *	m_curImage

std::unique_ptr< SharedImageSet >	m_outImages

std::unique_ptr< ImageResamplingAlgorithm >	m_resamplingAlgo

std::string	m_statsFile

SharedImageSet *	m_images

SharedImageSet *	m_labels

std::shared_ptr< SharedImage >	m_probabilityMap
	Probability map.

std::vector< vec3 >	m_detectedBone
	Detection result.

std::vector< vec3 >	m_segmentedBone
	Segmentation result.

float	m_curConfidenceScore

std::vector< double >	m_curBoneClassification

bool	m_useLegacyAlgorithm
	Use previous algorithm.

double	m_workingResolution
	Spacing for the working images.

int	m_frameStep
	Step between predicted frames.

int	m_nbSplinePoints
	Number of control points in the final spline.

std::string	m_predictionModelPath

std::unique_ptr< ML::MachineLearningModel >	m_predictionModel

float	m_detectionProbabilityThreshold
	Threshold on random forest probability for detection heuristic.

float	m_detectionDepthPenalization

double	m_marginAroundDetection
	Margin around detection for the mask in mm.

double	m_maskSmoothingScale
	Scale for the laplacian in mm.

double	m_marginAroundPrevious
	Margin around previous segmentation for the mask in mm.

float	m_segmentationThreshold
	Threshold for segmentation extraction.

float	m_segmentationGradientFactor

double	m_segmentationOffset

int	m_segmentationSmoothHalfSize

float	m_removeSidePercent
	Total percent of the points to be removed from the sides of each bone (0 1)

double	m_qualityThreshold
	Frames having lower tracking quality than this threshold are ignored.

bool	m_medianQualityThreshold
	If true, median tracking quality of all frames is used as the threshold.

Protected Attributes inherited from Algorithm
std::string	m_name
	Algorithm name.

Progress *	m_progress = nullptr
	Non-owing pointer to a progress interface. May be a nullptr.

FactoryInfo	m_factoryInfo = {}
	Record describing how this algorithm was instantiated by the AlgorithmFactory.

int	m_status = Status::Unknown
	Algorithm status after last call to compute()

std::vector< Action >	m_actions
	Map of key given by the id of the action, of the available actions of this algorithm.

Protected Attributes inherited from Configurable
std::vector< Param >	m_params
	List of all registered Parameter and SubProperty instances.

Member Function Documentation

◆ configure()

void configure ( const Properties * p )

overridevirtual

Configure this object instance by de-serializing the given Properties.

The default implementation will do so automatically for all registered Parameter and SubProperty instances.

See also: configuration() for the inverse functionality

Reimplemented from Configurable.

◆ configuration()

void configuration ( Properties * p ) const

overridevirtual

Serialize the current object configuration into the given Properties object.

The default implementation will do so automatically for all registered Parameter and SubProperty instances.

See also: configure() for the inverse functionality

Reimplemented from Configurable.

◆ compute()

void compute ( )

overridevirtual

Computes bone segmentation and prepares it for output.

Performs automatic bone detection on each sweep. If a bone is found, it then computes the segmentation and estimates outliers for it. If the parameter "Export Point Cloud" is set to true, it also exports the point cloud.

Note: Notifies listeners about the output change.

Implements Algorithm.

◆ takeOutput()

OwningDataList takeOutput ( )

overridevirtual

Get the calculated results.

Returns

OwningDataList filled in with up to 3 Data entities:

If the parameter "Intermediate Output" is set, includes the SweepDetectionResults instance for the current sweep.
If the compounded probability map is available (i.e., computeCompoundedProbability() was called successfully), includes a std::unique_ptr to the shared image set with the map.
Includes a std::vector of std::unique_ptr<PointCloud> representing the point clouds.

Reimplemented from Algorithm.

◆ sweep()

UltrasoundSweep * sweep ( ) const

inline

Returns: A pointer to the ultrasound sweep input image.

◆ bestSlice()

int bestSlice ( UltrasoundSweep * sweep ) const

inline

Returns: the index of the best slice for the requested sweep.

Parameters

sweep - sweep to look for the best slice for. If equal to nullptr then return the best slice for the current internal sweep.

◆ isResultAvailable()

bool isResultAvailable	(	int	slice,
		UltrasoundSweep *	sweep ) const

Returns: true if a segmented bone exists for the specified slice in the given sweep.

Parameters

slice	- the index of a slice to look for.
sweep	- the sweep to look in. If set nullptr then will look in the currently set internal sweep.

◆ detectedBone()

const std::vector< vec3 > & detectedBone	(	int	slice,
		UltrasoundSweep *	sweep ) const

inline

Returns the list of detected bone centerline points for a specific ultrasound slice.

This function provides the 3D points that represent the detected centerline of the bone within a given slice of an ultrasound sweep. If the sweep parameter is nullptr, the currently active internal sweep is used.

Returns: const reference to a std::vector of vec3 points representing the detected bone centerline

Parameters

slice	- index of the ultrasound slice.
sweep	- pointer to the UltrasoundSweep containing the slice. If nullptr, the internal sweep is used.

◆ segmentedBone()

const std::vector< vec3 > & segmentedBone	(	int	slice,
		UltrasoundSweep *	sweep ) const

inline

Returns the segmented bone centerline points for a specific ultrasound slice.

This function retrieves the 3D points representing the segmented bone centerline. for a given slice of an ultrasound sweep. If the sweep parameter is nullptr, the currently active internal sweep is used.

Returns: const reference to a std::vector of vec3 points representing the segmented bone centerline.

Parameters

slice	Index of the ultrasound slice.
sweep	Pointer to the UltrasoundSweep containing the slice. If nullptr, the internal sweep is used.

◆ outlierness()

const std::vector< double > & outlierness ( UltrasoundSweep * sweep ) const

inline

Returns: the std::vector of outlierness for the slices in the provided sweep.

Parameters

sweep - the sweep to look in. If set nullptr then will look in the currently set internal sweep.

◆ detectionConfidence()

const std::vector< double > & detectionConfidence ( UltrasoundSweep * sweep ) const

inline

Returns: the std::vector of confidence scores for each slice in the provided sweep.

Parameters

sweep - the sweep to look in. If set nullptr then will look in the currently set internal sweep.

◆ boneClassification()

const std::vector< std::vector< double > > & boneClassification ( UltrasoundSweep * sweep ) const

inline

Returns: the std::vector bone classification for each slice in the provided sweep.

Parameters

sweep - the sweep to look in. If set nullptr then will look in the currently set internal sweep.

◆ init()

void init ( )

overrideprotectedvirtual

Clear/resize all vectors of images/results and initialize the algorithm (e.g. load the ML model)

Reimplemented from BoneSegmentationBaseAlgorithm.

The documentation for this class was generated from the following file:

ImFusion/US/BoneSegmentationAlgorithm.h

Detailed Description

Classes

Public Member Functions

Static Public Member Functions

Protected Member Functions

Input

Output

Configurable Parameters

Outliers

Additional Inherited Members

Member Function Documentation

◆ configure()

◆ configuration()

◆ compute()

◆ takeOutput()

◆ sweep()

◆ bestSlice()

◆ isResultAvailable()

◆ detectedBone()

◆ segmentedBone()

◆ outlierness()

◆ detectionConfidence()

◆ boneClassification()

◆ init()