 |
ImFusion C++ SDK 4.5.0
|
|
ImFusion C++ SDK 4.5.0
|
#include <CranialPlugin/include/ImFusion/Cranial/TumorSegmentationAlgorithm.h>
Algorithm for segmenting tumour in one or several MR images. More...
Algorithm for segmenting tumour in one or several MR images.
The input can contain one to four images in the following modalities [T1, T1CE, T2, FLAIR] (see MRSequenceClassificationAlgorithm). The input images can be provided in any order since a modality classification will be performed by the algorithm. The input images are assumed to be pre-registered. The input images will be internally skull stripped by the algorithm. Images may have different resolutions and slice thicknesses. The output label map matches the first input image's resolution and slice thickness.
Public Types | |
| enum class | TumorRegion { NecroticTumour = 1 , Edema = 2 , EnhancingTumour = 3 } |
| Subset of regions to be segmented by the tumour segmentation algorithm. | |
| Public Types inherited from ImFusion::Algorithm | |
| enum | Status { Unknown = -1 , Success = 0 , Error = 1 , InvalidInput , IncompleteInput , OutOfMemoryHost , OutOfMemoryGPU , UnsupportedGPU , UnknownAction , AbortedByUser , User = 1000 } |
| Status codes. More... | |
Public Member Functions | |
| TumorSegmentationAlgorithm (std::vector< SharedImageSet * > input) | |
| const SharedImageSet * | tumorLabelMap () const |
| Returns a non-owning pointer to the generated tumor label map, or nullptr if the algorithm was not successful or if the output was already fetched via takeOutput(). | |
| const SharedImageSet * | ensembleSoftmaxMap () const |
| Returns a non-owning pointer to the ensemble softmax map, or nullptr if not requested. | |
| const SharedImageSet * | ensembleConfidenceMap () const |
| Returns a non-owning pointer to the ensemble confidence map, or nullptr if not requested. | |
| const SharedImageSet * | outputVariantLabelMaps () const |
| Returns a non-owning pointer to the variant label maps, or nullptr if not available. | |
| const std::map< uint8_t, float > & | classConfidenceScores () const |
| Mean ensemble confidence per tumor label class (empty if p_estimateConfidenceScore is false). | |
| Public Member Functions inherited from ImFusion::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 ImFusion::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 |
| Public Member Functions inherited from ImFusion::SignalReceiver | |
| SignalReceiver ()=default | |
| Default constructor. | |
| SignalReceiver (const SignalReceiver &other) | |
| Copy constructor, does not copy any existing signal connections from other. | |
| SignalReceiver & | operator= (SignalReceiver rhs) |
| Assignment operator, disconnects all existing connections, does not copy any existing signal connections from rhs. | |
| virtual | ~SignalReceiver () |
| Virtual destructor disconnects from all connected signals. | |
Public Attributes | |
| Parameter< std::string > | p_modelSet = {"modelSet", Models::Latest::ModelSetName, this} |
| Model set to use for all models (tumor segmentation and internal sub-models). | |
| Parameter< std::string > | p_modelSetVersion = {"modelSetVersion", Models::Latest::ModelSetVersion, this} |
| Model set version used for all models. | |
| Parameter< bool > | p_estimateConfidenceScore = {"estimateConfidenceScore", false, this} |
| Whether to compute a confidence score based on the model's confidence. | |
| Parameter< std::vector< bool > > | p_contrastDropout = {"contrastDropout", std::vector<bool>{true, true, false, true}, this} |
| Which contrast modalities (FLAIR, T1, T1CE, T2) to include in dropout ensembling. | |
| Parameter< std::vector< double > > | p_contrastWeights = {"contrastWeights", std::vector<double>{0.25, 0.25, 0.25, 0.25}, this} |
| Weights for each dropout variant (FLAIR, T1, T1CE, T2). | |
| Public Attributes inherited from ImFusion::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 ImFusion::Configurable | |
| Signal | signalParametersChanged |
| Emitted whenever one of the registered Parameters' or SubPropertys' signalValueChanged signal was emitted. | |
| void | compute () override |
| Execute the algorithm. | |
| 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. | |
| OwningDataList | takeOutput () override |
| Returns the label map. | |
| static bool | createCompatible (const DataList &data, Algorithm **a=0) |
| Standard methods for the Algorithm interface. | |
Additional Inherited Members | |
| Static Public Member Functions inherited from ImFusion::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 inherited from ImFusion::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. | |
| Protected Member Functions inherited from ImFusion::SignalReceiver | |
| void | disconnectAll () |
| Disconnects all existing connections. | |
| Protected Attributes inherited from ImFusion::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 ImFusion::Configurable | |
| std::vector< Param > | m_params |
| List of all registered Parameter and SubProperty instances. | |
|
overridevirtual |
Execute the algorithm.
Implements ImFusion::Algorithm.
|
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.
Reimplemented from ImFusion::Configurable.
|
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.
Reimplemented from ImFusion::Configurable.
|
overridevirtual |
Returns the label map.
Reimplemented from ImFusion::Algorithm.
|
inline |
Returns a non-owning pointer to the generated tumor label map, or nullptr if the algorithm was not successful or if the output was already fetched via takeOutput().
The label values represent the different regions represented by TumorRegion.
|
inline |
Returns a non-owning pointer to the ensemble softmax map, or nullptr if not requested.
Multi-channel float32: weighted average of softmax outputs from all dropout runs (one channel per class). Note: excluded from takeOutput() due to size - access via this accessor if needed.
|
inline |
Returns a non-owning pointer to the ensemble confidence map, or nullptr if not requested.
Single-channel image: per-voxel softmax amplitude looked up at the ArgMax of the ALL prediction.
|
inline |
Returns a non-owning pointer to the variant label maps, or nullptr if not available.
Multi-frame SharedImageSet: one label map per executed dropout variant, in the order they were run (FLAIR, T1, T1CE, T2). The full-contrast prediction is excluded and can be accessed via tumorLabelMap(). Returns nullptr when p_estimateConfidenceScore is false, or when only one input image is provided (a single contrast cannot produce any variants). The output Properties contains VariantNames and VariantWeights in the same order as frames.
|
inline |
Mean ensemble confidence per tumor label class (empty if p_estimateConfidenceScore is false).
Maps label value to mean per-voxel confidence in [0, 1]. Only classes present in the output label map have entries; absent classes are not included.
| Parameter<bool> ImFusion::Cranial::TumorSegmentationAlgorithm::p_estimateConfidenceScore = {"estimateConfidenceScore", false, this} |
Whether to compute a confidence score based on the model's confidence.
When enabled, the algorithm outputs both the label map and a confidence score image. If contrast dropout is enabled, multiple forward passes are run with different contrasts dropped, and the softmax outputs are ensembled.
| Parameter<std::vector<bool> > ImFusion::Cranial::TumorSegmentationAlgorithm::p_contrastDropout = {"contrastDropout", std::vector<bool>{true, true, false, true}, this} |
Which contrast modalities (FLAIR, T1, T1CE, T2) to include in dropout ensembling.
When an entry is false, that contrast is never dropped and is always present in all variants. Useful to pin the most informative contrast (e.g., T1CE) while still dropping others.
| Parameter<std::vector<double> > ImFusion::Cranial::TumorSegmentationAlgorithm::p_contrastWeights = {"contrastWeights", std::vector<double>{0.25, 0.25, 0.25, 0.25}, this} |
Weights for each dropout variant (FLAIR, T1, T1CE, T2).
When computing softmax confidence, the model runs once with all contrasts (weight 1.0), then once per enabled dropout contrast (weight = this vector's value). The softmax outputs are averaged using these weights (normalized).