temporally_related_to
Related in time by preceding, co-occuring with, or following
follows
precedes
has_property
has_pacs_study
applied_to
To make use of something vs an entity
computed_using
Refers to the process / tool used to produce the computed entity
converted_to
Transformed into something
contains
to have within
defined_by
Which specifies particular properties
delineation_of
Obtained by delineate a contour
has_version_status
developed_by
Created an entity (e.g. a software)
develops
Act to create something new
direction_of
extracted_from
has_basis_function
Specifies basis function of a filter
has_calculation_run
Related to a computational process (calculation run)
has_computed
Has produced as result a computation
has_delineation
has development_status
Specifies what is the status of the development of the SW from first developed test versions to live (in a production environment) versions
has_direction
Specifies the direction / orientation of a filter
has_distance
Specifies the distance (e.g. norm)
has_filter
Specifies the imaging filter
has_function
Speficies the mathematical function
has_GreyLevelRound
Specifies the grey level round
has_intensity_fraction
Specifies the intensity fraction (refers to class IntensityFraction) used to compute the volume at intensity fraction
has_label
has_license
The relationship between an entity and the set of legal restrictions, i.e. license, which are applied in using or otherwise interacting with that entity. Eg. relationship between software and a software license
has_max
Specifies maximum value
has_method
Specifies computational method
has_min
Specifies minimum value
has_orientation
Specifies orientation along the space
has_PartialVolumeCutOff
Specifies Partial Volume cut off value
has_processing
Specifies any post processing applied to an image
has_programming_language
This predicate is used to describe which is the programming language in which a SW or application was written
has_radiomics_feature
Specifies the radiomic feature related to the entity
has_scale
Speficies if a particular scale (e.g. logaritmic has been applied)
has_segmentation_method
This predicates is used to link a Region of Interest with the method used to generate it. It can be both a manual delineation or a SW based the delineation
has_unit
Specifies the unity associated to a quantity
has_version
This predicates is use to link the software with the name of the version. Usually the version name is represented by a number, which represents the release of the SW
has_volume_fraction
The predicate is used to specify the volume fraction (refers to class VolumeFraction) used to compute the intensity at volume fraction (refers to class IntensityAtVolumeFraction) feature
has_voxel_dimensionX
has_voxel_dimensionY
has_voxel_dimensionZ
implemented_in
To put into effect according to or by means of a definite procedure
is_connected_to
is_label_of
is_part_of
is_processing_of
is_programming_language_of
has_featurespace
define feature space
has_imagespace
define image space
has_imaging_modality
has_discretization_algorithm
is_radiomics_feature_of
performs
is_version_of
performed_by
has_FixedBinSize
originated_from
Created from a certain context
segmented
has_interpolation_method
has_equalisation
has_range
has_FixedBinNumber
has_outlier_removal
is_FixedBinSize_of
is_outlier_removal_of
run_on
http://purl.bioontology.org/ontology/NCIT
GrossTumourVolume
The Gross Tumor Volume (GTV) is the gross demonstrable extent and location of the tumor. The GTV may consist of a primary tumor (primary tumor GTV or GTV-T), metastatic regional node(s) (nodal GTV or GTV-N), or distant metastasis (metastatic GTV, or GTV-M). Typically, different GTVs are defined for the primary tumor and the regional node(s). But in some particular clinical situations, it might well be that the metastatic node cannot be distinguished from the primary tumor, e.g., a nasopharyngeal undifferentiated carcinoma infiltrating posterolaterally into the retropharyngeal space, including possible infiltrated nodes. In such situations, a single GTV encompassing both the primary tumor and the node(s) may be delineated. (ICRU 83)
For more information: http://dx.doi.org/10.1093/jicru/ndq001
http://purl.bioontology.org/ontology/NCIT
MagneticResonanceImaging
http://purl.bioontology.org/ontology/NCIT
Patient
A person who receives medical attention, care, or treatment, or who is registered with medical professional or institution with the purpose to receive medical care when necessary
http://purl.bioontology.org/ontology/NCIT
PositronEmissionTomography
http://purl.bioontology.org/ontology/NCIT
ComputedTomography
http://purl.bioontology.org/ontology/NCIT
Scan
The data or image obtained by gathering information with a sensing device
http://purl.bioontology.org/ontology/NCIT
MedicalImage
Any record of a medical imaging event whether physical or electronic
http://purl.bioontology.org/ontology/NCIT
Person
A human being
Scale
A measurement that uses a mathematical transformation of a physical quantity instead of the quantity itself
http://purl.bioontology.org/ontology/NCIT
ImagingRegionOfInterest
A specific area of interest defined by a sequence of image overlays or a sequence of contours described as a single point (for a point ROI) or more than one point (representing and open or closed polygon)
http://purl.bioontology.org/ontology/STY
Event
A broad type for grouping activities, processes and states
http://purl.bioontology.org/ontology/STY
Entity
A broad type for grouping physical and conceptual entities
http://purl.bioontology.org/ontology/STY
PhysicalObject
An object perceptible to the sense of vision or touch or than can be used by an human
http://purl.bioontology.org/ontology/STY
ConceptualEntity
A broad type for grouping abstract entities or concepts
http://purl.bioontology.org/ontology/STY
QuantitativeConcept
A concept which involves the dimensions, quantity or capacity of something using some unit of measure, or which involves the quantitative comparison of entities
http://purl.bioontology.org/ontology/STY
FunctionalConcept
A concept which is of interest because it pertains to the carrying out of a process or activity
ProgrammingLanguage
A language in which source code is written, intended to executed/run by a software interpreter. Programming languages are ways to write instructions that specify what to do, and sometimes, how to do it
http://purl.bioontology.org/ontology/UO
Unit
A unit of measurement is a standardized quantity of a physical quality
http://purl.bioontology.org/ontology/UO
LengthUnit
A unit which is a standard measure of the distance between two points
http://purl.bioontology.org/ontology/UO
Meter
A length unit which is equal to the length of the path traveled by light in vacuum during a time interval of 1/299 792 458 of a second
http://purl.bioontology.org/ontology/UO
Centimeter
A length unit which is equal to one hundredth of a meter or 10^[-2] m
http://purl.bioontology.org/ontology/UO
Millimeter
A length unit which is equal to one thousandth of a meter or 10^[-3] m
http://purl.bioontology.org/ontology/UO
AreaUnit
A unit which is a standard measure of the amount of a 2-dimensional flat surface
http://purl.bioontology.org/ontology/UO
SquareMeter
An area unit which is equal to an area enclosed by a square with sides each 1 meter long
http://purl.bioontology.org/ontology/UO
SquareCentimeter
An area unit which is equal to one thousand of square meter or 10^[-3] m^[2]
http://purl.bioontology.org/ontology/UO
SquareMillimeter
An area unit which is equal to one millionth of a square meter or 10^[-6] m^[2]
http://purl.bioontology.org/ontology/UO
VolumeUnit
A unit which is a standard measure of the amount of space occupied by any substance, whether solid, liquid, or gas
http://purl.bioontology.org/ontology/UO
CubicMeter
A volume unit which is equal to the volume of a cube with edges one meter in length. One cubic meter equals to 1000 liters
http://purl.bioontology.org/ontology/UO
CubicCentimeter
A volume unit which is equal to one millionth of a cubic meter or 10^[-6] m^[3], or to 1 ml
FrequencyUnit
A unit which is a standard measure of the number of repetitive actions in a particular time
Hertz
A frequency unit which is equal to 1 complete cycle of a recurring phenomenon in 1 second
AngleUnit
A unit which is a standard measure of the figure or space formed by the junction of two lines or planes.
Radian
A plane angle unit which is equal to the angle subtended at the center of a circle by an arc equal in length to the radius of the circle, approximately 57 degrees 17 minutes and 44.6 seconds.
Degree
A plane angle unit which is equal to 1/360 of a full rotation or 1.7453310^[-2] rad.
http://purl.bioontology.org/ontology/SEDI
RTDose
http://purl.bioontology.org/ontology/SEDI
RTPlan
http://purl.bioontology.org/ontology/SEDI
RTStructureSet
http://purl.bioontology.org/ontology/ROO
GTV(ROI)
A region of interest based on a delineation of the Gross Tumor Volume. The label adheres to the standardized naming convention proposed by Santanam et al (http://dx.doi.org/10.1016/j.ijrobp.2011.09.054)
http://purl.bioontology.org/ontology/ROO
PTV(ROI)
A region of interest based on a delineation of the Planning Target Volume. The label adheres to the standardized naming convention proposed by Santanam et al (http://dx.doi.org/10.1016/j.ijrobp.2011.09.054)
http://purl.bioontology.org/ontology/ROO
RadiationOncologyRegionOfInterest
The class of regions of interest used in Radiation Oncology and adhering to the Standardized Naming Conventions in Radiation Oncology
http://purl.bioontology.org/ontology/ROO
TargetVolume(ROI)
A region of interest based on a delineation of a Target Volume
http://purl.bioontology.org/ontology/ROO
OrganAtRisk(ROI)
A region of interest based on a delineation of an Organ-At-Risk (OAR)
http://purl.bioontology.org/ontology/ROO
RadiationOncologyDICOMFunctionalConcept
A concept which is of interest because it is related to DICOM objects used in radiation oncology
http://purl.bioontology.org/ontology/ROO
CTV(ROI)
A region of interest based on a delineation of the Clinical Target Volume. The label adheres to the standardized naming convention proposed by Santanam et al (http://dx.doi.org/10.1016/j.ijrobp.2011.09.054)
http://purl.bioontology.org/ontology/ROO
RadiationOncologyFunctionalConcept
A concept which is of interest because it pertains to the carrying out of radiation oncology.
Examples are target and other irradiated volumes, margins etc
http://purl.bioontology.org/ontology/ROO
PlanningTargetVolume
The Planning Target Volume (PTV) is a geometrical concept introduced for treatment planning and evaluation. It is the recommended tool to shape absorbed-dose distributions to ensure that the prescribed absorbed dose will actually be delivered to all parts of the CTV with a clinically acceptable probability, despite geometrical uncertainties such as organ motion and setup variations. It is also used for absorbed-dose prescription and reporting. It surrounds the representation of the CTV with a margin such that the planned absorbed dose is delivered to the CTV. This margin takes into account both the internal and the setup uncertainties. The setup margin accounts specifically for uncertainties in patient positioning and alignment of the therapeutic beams during the treatment planning, and through all treatment sessions. (ICRU 83)
For more information: http://dx.doi.org/10.1093/jicru/ndq001
http://purl.bioontology.org/ontology/ROO
TargetVolume
A Target Volume is a volume of tissue or of a geometrical concept forming the target for irradiation during radiation oncology
http://purl.bioontology.org/ontology/ROO
OrganAtRisk
The Organ-At-Risk (OAR) or critical normal structures are tissues that if irradiated could suffer significant morbidity and thus might influence the treatment planning and/or the absorbed-dose prescription. (ICRU 83)
For more information: http://dx.doi.org/10.1093/jicru/ndq001
Software
DevelopmentStatus
Development status is an information content entity which indicates the maturity of a sofrware entity within the context of the software life cycle.
Alpha
Alpha is a development status which is applied to software by the developer/publisher during initial development and testing. Software designated alpha is commonly unstable and prone to crashing. It may or may not be released publicly
Beta
Beta is a development status which is generally applied to software by the developer/publisher once the majority of features have been implemented, but when the software may still contain bugs or cause crashes or data loss. Software designated beta is often released publicly, either on a general release or to a specific subset of users called beta testers.
Live
Live is a development status which is applied to software that has been designated as suitable for production environments by the developer/publisher. If a non-free product, software at this stage is available for purchase
Obsolete
Sofware is no longer being supplied by the developers/publishers
Mantained
Software has developers actively maintaining it (fixing bugs)
IBSI
GLDZM_LargeDistanceLowGreyLevelEmphasis
This feature emphasises runs in the upper right quadrant of the GLDZM, where large zone distances and low grey levels are located.
IBSI
GLCM_Contrast
Contrast assesses grey level variations. It is defines as in https://doi.org/10.5589/m02-004
IBSI
GLCM_ClusterProminence
The cluster prominence is a measure of asymmetry. When the cluster prominence value is high, the image is less symmetric
BSplineInterpolation
This filter performs interpolation using a B-spline of order n
IBSI
IntensityHistogramMode
Most common discretised grey level present in the histogram
ExponentialDecay
Defined as exp(-x)
IBSI
VolumeAtIntensityFraction
It is the largest volume fraction that has an intensity fraction of at least a certain percentage. See also https://doi.org/10.1016/j.patcog.2008.08.011
Scaling
This class includes all the filters which are used to resize an image. They can be used both to increase or reduce the dimensions
IBSI
IntensityHistogramUniformity
The uniformity is a measure of the randomness of the grey levels distirbution histogram
IBSI
Compactness2
Compactness 2 is anothere measure to describe how sphere-like the volume is
IBSI
AreaDensityMVEE
The surface density minimum volume enclosing ellipsoid is defined as the ratio between the area and the surface of the eilpsoide as defined in feature VolumeDensityMVEE
2DAverage
Averaged over slices and directions
IBSI
GLSZM_GreyLevelVariance
This feature estimates the variance in zone counts for the grey levels
IBSI
SurfaceArea
The surface area is calculated from the ROI mesh, by summing over the face area surfaces
IBSI
IntensityHistogramKurtosis
Kurtosis is a measure of how outlier-prone a distribution is. The kurtosis of the normal distribution is 3. Distributions that are more outlier-prone than the normal distribution have kurtosis greater than 3; distributions that are less outlier-prone have kurtosis less than 3
SquareHU
IBSI
RobustMeanAbsoluteDeviation
The mean absolute deviation feature may be influenced by outliers. To increase robustness, the set of grey levels can be restricted to those which lie closer to the center of the distribution.
IBSI
NGLDM_DependenceCountEnergy
Defined as second moment in Sun and Wee (1983)
BellInterpolation
Bell uses a kernel to interpolate the pixels of the input image. The kernel is defined as:
0.75-|x| if x<0.5
0.5 (|x| - 1.5)^2 if 0.5 < x < 1.5
0 otherwise
IBSI
IntensityHistogramVariance
Variance of the intensities
FeatureParameterSpace
The feature parameter space is defined as the space that includes all the possible radiomics computation methods that can be applied to certain radiomics features. Subclasses are: aggregation re-segmentation, interpolation, features specific parameters, discretization
LHH
CubicMillimeterHU
LaplacianFilter
Convolutes with a laplacian kernel the original image
IBSI
IntensityAtVolumeFractionDifference
The difference between grey levels at two different fractional volumes
MarchingCubes
GaussianFilter
This filter applies a gaussian kernel with a pre-defined average value and variance to each pixel of an image It is used to blur image and remove details and noise
IBSI
IntensityHistogramCoefficientOfVariation
The intensity Histogram Coefficient of Variation measures the dispersion of the histogram
IBSI
IntensityHistogramMeanAbsoluteDeviation
Measure of dispersion from the mean of the histogram
IBSI
GLCM_DifferenceVariance
The variance for the diagonal probabilities
Voxel
IBSI
VolumeAtIntensityFractionDifference
The difference between the volume fractions at two different intensity fractions. See also https://doi.org/10.1016/j.patcog.2008.08.011
IBSI
GLCM_ClusterTendency
The cluster tendency indicates into how many clusters the gray levels present in the image can be classified
ArcSin
Inverse Sine
IBSI
GLDZM_SmallDistanceHighGreyLevelEmphasis
This feature emphasises runs in the lower left quadrant of the GLDZM, where small zone distances and high grey levels are located
IBSI
NGLDM_DependenceCountVariance
This feature estimates the variance in dependence counts for the different dependence counts possible
IBSI
GLCM_InverseVariance
The inverse variance feature measures how the gray tone differences are distributed in pair elements
IBSI
Variance
The variance from grey level distribution
I10minusI90
Difference between the intensity at volume fraction 10 and intensity at volume fraction 90
ArcCos
Inverse Cosine
SinH
Hyperbolic Sine
IBSI
NGLDM_LowDependenceLowGreyLevelEmphasis
This feature emphasises neighbouring grey level dependence counts in the upper left quad- rant of the NGLDM, where low dependence counts and low grey levels are located
NearestNeighbour
Nearest neighbour interpolation assigns grey levels in the output grid to the values of the closest voxels in the input grid.
ColliageFilter
IBSI
NGLDM_DependenceCountEntropy
The entropy for the dependence counts
MorphologicalSpecificParameters
IBSI
NGLDM_GreyLevelNonUniformity
This feature assesses the distribution of neighbouring grey level dependence counts over the grey values. The feature value is low when dependence counts are equally distributed along grey levels
Haar
Cosine
Cosine Trigonometric Function
IBSI
GLRLM_HighGreyLevelRunEmphasis
The HGLRE measures the distribution of high gray level values. The HGRE is high for the image with highgray level values
IsoValue
Iso-value algorithm with isolevels
ImageVolume_GreyLevelRound
EuclideanNorm
The "ordinary" straight-line distance between two points in Euclidean space
IBSI
GLDZM_ZoneDistanceEntropy
Entropy for the zone distances
IBSI
IntensityAtVolumeFraction
Minimum grey level present in at most a certain percentage of the volume. See also https://doi.org/10.1016/j.patcog.2008.08.011
IBSI
GLRLM_ShortRunHighGreyLevelEmphasis
The SRHGLE measures the joint distribution of short runs and high gray level values. The SRHGE is high for the image with many short runs and high gray level values
Pyradiomics
IntensityHistogramStandardDeviation
The Standard Deviation measures the amount of variation or dispersion from the Mean Value
CalculationRunSpace
Properties related to the process used to perform a calculation (e.g. a SW processing something)
IBSI
IntensityHistogramPercentile10
10 Percentile of the histogram
IBSI
GLSZM_ZoneSizeEntropy
Entropy related to the zone sizes
IBSI
GLCM_JointMaximum
Probability corresponding to the most common grey level co-occurence in the GCLM
Closing
This filter performs morphological closing. The morphological operation includes an a dilation followed by an erosion, using the same structuring element for both operations
Morphological
Morphological features describe geometric aspects of a region of interest (ROI), such as area and volume. Morphological features are based on ROI voxel representations of the volume.
Distance
The space separating two objects or points.
IBSI
NGTDM_Complexity
Complex textures are non-uniform and rapid changes in grey levels are common
IBSI
GLRLM_RunEntropy
IBSI
GLRLM_ShortRunLowGreyLevelEmphasis
The SRLGLE measures the joint distribution of short runs and low gray level values. The SRLGE is high for the image with many short runs and lower gray level values
IBSI
GLSZM_SmallZoneHighGreyLevelEmphasis
This feature emphasises runs in the lower left quadrant of the GLSZM, where small zone sizes and high grey levels are located
WaveletDirection
IBSI
GLDZM_ZoneDistanceNonUniformityNormalised
This is the normalised version of the zone distance non-uniformity feature
3DMerging
Merged 3D directions
IBSI
GLCM_InverseDifference
Inverse difference is a measure of homogeneity. Grey level co-occurrences with a large difference in levels are weighed less, thus lowering the total feature score. The feature score is maximal if all grey levels are the same
IBSI
GLRLM_RunLengthNonUniformityNormalised
This is the normalised version of the run length non-uniformity feature
NGLDMSpecificParameters
IBSI
NGLDM_HighDependenceEmphasis
This feature emphasises high neighbouring grey level dependence counts
NeighbourhoodGreyToneDifferenceMatrix
The neighbourhood grey tone difference matrix (NGTDM) contains the sum of grey level differences of pixels/voxels with discretised grey level i and the average discretised grey level of neigh- bouring pixels/voxels within a distance d. See also https://doi.org/10.1109/21.44046
IBSI
Kurtosis
Kurtosis is a measure of how outlier-prone a distribution is. The kurtosis of the normal distribution is 3. Distributions that are more outlier-prone than the normal distribution have kurtosis greater than 3; distributions that are less outlier-prone have kurtosis less than 3
IBSI
AreaDensityOMBB
The area density oriented bounding box is the ratio between the area and the surface area of the same bounding box as calculated for the VolumeDensityOMBB
3DAverage
Averaged over 3D directions
MeanFilter
The mean filter works on a n x n subregion of the image. At each position the center pixel is replaced by the mean value
IBSI
GLRLM_LongRunLowGreyLevelEmphasis
This feature emphasises runs in the upper right quadrant of the GLRLM, where long run lengths and low grey levels are located
EdgeEnhancement
This filter detects edges in different orientation and enhances them working on pixels with different orientations
IBSI
GLSZM_LargeZoneHighGreyLevelEmphasis
This feature emphasises runs in the lower right quadrant of the GLSZM, where large zone sizes and high grey levels are located
IBSI
NGLDM_LowDependenceHighGreyLevelEmphasis
This feature emphasises neighbouring grey level dependence counts in the lower left quadrant of the NGLDM, where low dependence counts and high grey levels are located
Orientation
The orientation, angular position, or attitude of an object such as a line, plane or rigid body is part of the description of how it is placed in the space it is in
ImageFilterSpace
The image filter space is a container for all the filters that can be applied during the computation of a feature
2.5DAverage
Merged per direction and averaged
IBSI
GLCM_SecondMeasureOfInformationCorrelation
IBSI
GLSZM_GreyLevelNonUniformity
This feature assesses the distribution of zone counts over the grey values. The feature value is low when zone counts are equally distributed along grey levels
Average
This filter takes as input two images and produces an image which has as pixel values the average of the values of the starting images
Angular
Orientation referred to angular directions
FixedBinNumber
In the fixed bin number method, grey levels are discretised to a fixed number of bins. The number of bins should always be specified
IBSI
GLDZM_HighGreyLevelZoneEmphasis
The feature emphasises high grey levels
ImageOperations
This class is a container for most common image filters used to pre-process an image before feature extraction
VoxelDimensionX
Translation
This filter rotates an image according to a defined rotation matrix
IBSI
Skewness
The skewness measures the degree of histogram of gray levels asymmetry around the central value
WaveletProperties
IBSI
CentreOfMassShift
The distance between the ROI volume centroid and the intensity-weigthted ROI volume centroid measures the placement of high and low intensity regions within the volume
IBSI
GLDZM_LargeDistanceHighGreyLevelEmphasis
This feature emphasises runs in the lower right quadrant of the GLDZM, where large zone distances and high grey levels are located
3D
Calculated over the volume
IntVolHistSpecificParameters
IBSI
SphericalDisproportion
Spherical disproportion is a measure to describe how sphere-like the volume is
Automated
Segmentation is automated performed without requiring any additional interaction by the user. These tecnqiues usually make use of AI
ImageVolume
Volume derived from an image
IBSI
Maximum3DDiameter
The maximum 3D diameter is the distance between the two most distant vertices in the ROI mesh vertices sets
SegmentationMethods
This class is a container for all the main algorithms used to segment / generate a ROI
HHH
GreyLevelCoOccurenceMatrix
The grey level co-occurrence matrix (GLCM) is a matrix that expresses how combinations of discretised grey levels of neighbouring pixels, or voxels in a 3D volume, are distributed along one of the image directions. In a 3 dimensional approach to texture analysis, the direct neighbourhood of a voxel consists of the 26 directly neighbouring voxels. Thus, there are 13 unique direction vectors within a neighbourhood volume for distance
ManhattanNorm
The distance between two points is the sum of the absolute differences of their Cartesian coordinates.
ImagingModality
BiCubicInterpolation
The filter makes use of third degree polynomial function to interpolate two pixels
FilterProperties
Specifies propery of an image filter
VoxelDimension
VolumetricHU
IBSI
GLDZM_LargeDistanceEmphasis
This feature emphasises large distances
MedianFilter
The median filter works on a n x n subregion of the image. At each position the center voxel is replaced by the median value
FixedBinSizeEqual
FixedBinCountEqual
NGTDMSpecificParameters
IBSI
Flatness
The flatness is the ratio of the major and the least axis lengths. The flatness is expressed as an inverse ratio: 1 completely not flat; smaller values express objects which are increasingly flatter
IBSI
MoranIndex
Moran’s I index is an indicator of spatial autocorrelation. See also https://doi.org/10.1093/biomet/37.1-2.17
IBSI
MedianAbsoluteDeviation
Median absolute deviation is similar in concept to mean absolute deviation, but measures dispersion from the median instead of mean
IBSI
The energy is the square sum of all the gray levels associated to an image
Energy
IBSI
NGLDM_HighDependenceLowGreyLevelEmphasis
This feature emphasises neighbouring grey level dependence counts in the lower left quadrant of the NGLDM, where high dependence counts and low grey levels are located
IBSI
GLCM_InverseDifferenceNormalised
Normalized inverse difference as suggested in https://doi.org/10.5589/m02-004
IBSI
GLCM_Correlation
The correlation feature shows the linear dependence of gray level values in the cooccurence matrix
Gaussian
Defined as exp(-x^2)
Sine
Sine Trigonometric Function
IBSI
GearyMeasure
Geary’s C measures spatial autocorrelation. See also https://doi.org/10.2307%2F2986645
IBSI
NGTDM_Busyness
Textures with large changes in grey levels between neighbouring voxels are called busy
LLL
IBSI
GLCM_DifferenceEntropy
The entropy for the diagonal probabilities
CubicConvolution
Cubic convolution uses the same procedures as cubic spline (refers CubicSpline), but approximates the solution using a convolution filter.
SavitzkyGolayFilter
This filter is also called also called digital smoothing polynomial filters or least-squares smoothing filter. It is used to "smooth out" a noisy signal whose frequency span (without noise) is large. This filter minimizes the least-squares error in fitting a polynomial to frames of noisy data
GradientEdgeBased
This method makes use of the gradient operator (first order derivatives) to detect edges
HoughTransformBased
The purpose of the technique is to find imperfect instances of objects within a certain class of shapes by a voting procedure. This voting procedure is carried out in a parameter space, from which object candidates are obtained as local maxima in a so-called accumulator space that is explicitly constructed by the algorithm for computing the Hough transform
LaplacianEdgeBased
This method makes use of the laplacian operator (second order derivatives) to detect edges
MarrHildrethEdgeBased
The Marr–Hildreth edge detection method operates by convolving the image with the Laplacian of the Gaussian function, or, as a fast approximation by Difference of Gaussians. Then, zero crossings are detected in the filtered result to obtain the edges
ModelBased
This class includes all the methods for segmentation which are based on the fact the structure of organs has a repetitive form of geometry and can be modeled probabilistically for variation of shape and geometry
ActiveShape
The ASM model finds the main variations in the training data using Principal Component Analysis (PCA), which enables the model to automatically recognize if a contour is a possible/good object contour
AppereanceModel
An active appearance model (AAM) is a computer vision algorithm for matching a statistical model of appearance - a combination of shape and texture - to a new image
AtlasModel
The segmentation is performed trying to extract prior knowledge from a reference image often called atlas
DeformableModel
The implicit deformable models, also called implicit active contours or level sets, are designed to handle topological changes naturally
RegionBased
This class includes all the the methods for segmentation which are based on working on pixels / regions with similar intensities in order to try to group similar regions. Starting point of most of algorithms is the manual selection of seed points
RegionGrowing
The region growing looks for group of pixels with similar intensities. It starts with a group of pixels (seeds) belonging to the structure of interest. Neighboring pixels are examined one at a time and added to the growing region, if they are sufficiently similar. The procedure continues until no more pixels can be added
RegionSplitting
It is just opposite to region merging and whole image is continuously split until no further splitting of a region is possible
SplitAndMerge
This is the combination of splits and merges utilizing the advantage of the two methods. This method is based on quad quadrant tree representation of data whereby image segment is split into four quadrants provided the original segment is non-uniform in properties. After this the four neighboring squares are merged depending on the uniformity of the region (segments). This split and merge process is continued until no further split and merge is possible
WatershedAlgorithm
Watershed algoritm it Is a region-based tecnique that utilizes image morphology. It requires the selection of at least one marker (seed point). Once the objects are marked they can be grown using the watershed transformation. It is analogous to the notion of a catchment basin of a heightmap. In short, a drop of water following the gradient of an image flows along a path to finally reach a local minimum
GlobalThresholding
It is based on the assumption that the image has a bimodal histogram and, therefore, the object can be extracted from the background by a simple operation that compares image values with a threshold value T. The result of the process is a binary image
LocalAdaptiveThresholding
Local thresholding can be applied by: splitting an image into subimages and calculating thresholds for each subimage; examining the image intensities in the neighborhood of each pixel
TimeStamp
SemiAutomated
Segmentation of a ROI which includes a mimimum interaction by the user. E.g delination of contours or selection of starting seed points or the value of a threshold
EdgeBased
This class includes all the methods for segmentation which are based on marking of discontinuities in gray level, color etc., and often these edges represent boundaries between objects. The methods divide an image on the basis of its boundaries
RayCasting
Manual
Manual Segmentation of a ROI, usually performed delineating the contour of the region in each slice
InPolygon
PartialVolumeEffectCorrection
KMean
k-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean, serving as a prototype of the cluster
CMeansFuzzy
This algorithm works by assigning membership to each data point corresponding to each cluster center on the basis of distance between the cluster center and the data point. More the data is near to the cluster center more is its of distance between the cluster center and the data point
UnsupervisedMethods
In the unsupervised category we can put all the methods which does not require having labelled data because they are cluster based
SupportVectorMachine
A Support Vector Machine (SVM) is a discriminative classifier formally defined by a separating hyperplane. In the segmentation case the hyperplane is used to separate between foreground and background pixels
NeuralNetwork
The most famous class of ANN. The network needs to be trained on training data before being used. A list of features use by the net has to be defined
DeepLearning
SupervisedMethods
In the supervised category, we can place mostly Artificial Neural Network (ANN) algorithms. Every supervised methods needs a definition of a training set, but also as input a feature vector
IBSI
NGLDM_HighGreyLevelCountEmphasis
The feature emphasises high grey levels
Norm
ImageSpace
The image space contains information about an image
SUV
IBSI
GLCM_SumVariance
The variance for the cross-diagonal probabilities
ROIMask
Binary mask as result of a segmentation process
IBSI
NGLDM_DependenceCountNonUniformityNormalised
This is a normalised version of the dependence count non-uniformity feature
GaborFilter
This filter applies a Gabor filter with a specified wavelength (in pixels) and orientation (in degrees)
IBSI
GLRLM_GreyLevelNonUniformityNormalised
This is the normalised version of the grey level non-uniformity feature
FeatureSpecificParameters
Container for parameters and computation methods used to derive features
IBSI
IntensityHistogramPercentile90
90 percentile of the histogram
Institution
An organization founded for a religious, educational, professional, or social purpose
ManufacturedObject
A physical object made by human beings
Phantom
ManufacturedObjectProperties
Properties and characteristics related to any manufactered object
SoftwareProperties
All the properties used to define a SW
License
A software license is a legal instrument (usually by way of contract law, with or without printed material) governing the use or redistribution of software
OpenSource
Open source licenses are licenses that comply with the Open Source Definition — in brief, they allow software to be freely used, modified, and shared
Proprietary
Proprietary software is computer software for which the software's publisher or another person retains intellectual property rights—usually copyright of the source code, but sometimes patent rights
Version
IBSI
GLSZM_ZonePercentage
This feature assesses the fraction of the number of realised zones and the maximum num- ber of potential zones. Strongly linear or highly uniform ROI volumes produce a low zone percentage
Percentage
IBSI
GLCM_SumEntropy
The entropy for the cross-diagonal probabilities
IntensityVolumeHistogram
The intensity-volume histogram (IVH) of the voxel grey level of the ROI intensity mask describes the relationship between discretised grey level i and the fraction of the volume containing at least grey level $i$, $\nu$. See definition in https://doi.org/10.1016/j.patcog.2008.08.011
Frequency
Frequency used in a imaging filter
IBSI
MinorAxisLength
The minor axis length of the ROI provides a measure of how fare the volume extends along the second largest axis. The minor axis length is twice the semi-axis length, determined using the second largest eigenvalue obtained by PCA on the point of the voxel centers
IBSI
VolumeDensityAABB
Volume density is the fraction of the ROI volume and a comparison volume. This feature is also called extent
See https://doi.org/10.1016/j.patcog.2008.08.011
PostAcquisitionProcessing
Sum
This filter performs the sum, pixel by pixel, of two different images
ImageNonUniformityCorrection
Logaritmic
ChebyshevNorm
Chebyshev distance (or Tchebychev distance), maximum metric, is a metric defined on a vector space where the distance between two vectors is the greatest of their differences along any coordinate dimension.
LocalAreaHistoEqualization
This filter applies the same concept of the Histo Equalization filter, but to small, overlapping local areas of the image
IBSI
Elongation
Elongation is the ratio between the major and minor axis lengths. Elongation is espressed as inversed ratio: 1 means completely not elongated (e.g. a sphere). Smaller values express greater elongation of the ROI volume
FixedBinSize
A new bin is assigned every certain grey levels. The fixed bin size method has the advantage of maintaining a direct relationship with the original intensity scale, which could be useful for functional imaging modalities such as PET.
IBSI
Mean
The mean grey level from grey level distribution
Opening
This filter performs morphological opening. The morphological opening is composed by an erosion followed by an erosion, using the same structuring element for both operations
SquareSUV
IBSI
NGTDM_Coarseness
Grey level differences in coarse textures are generally small due to large-scale patterns. Summing differences gives an indication of the level of the spatial rate of change in intensity
IBSI
Sphericity
Sphericity is a further measure to describe how sphere-like the volume is
IBSI
Percentile10
The statistical 10th percentile of the grey level distribution
IBSI
GLDZM_GreyLevelVariance
This feature estimates the variance in zone counts for the grey levels
ReSegmentationParameters
Re-segmentation entails updating the ROI mask R based on certain corresponding voxel intensities. Re-segmentation may be performed to exclude voxels from a previously segmented ROI, and is performed after interpolation.
Biorthogonal3.5
IBSI
GLCM_Autocorrelation
The autocorrelation compares all possible pixel pairs and reporting the likelihood that both will be bright as a function of the distance and direction of separation
IBSI
VolumeDensityCH
The volume density convex hull is defined as the ration between the volume and the volume of the ROI mesh convex hull set. This feature is also called solidity. See also https://dx.doi.org/10.1016%2Fj.patcog.2008.08.011
IBSI
AreaDensityAABB
Area Density considers the ratio of the ROI surface area and the surface area of the axis-aligned bounding box enclosing the ROI mesh vertex set. See also https://doi.org/10.1016/j.radonc.2016.07.007
IBSI
GLRLM_GreyLevelNonUniformity
This feature assesses the distribution of runs over the grey values (Galloway, 1975). The feature value is low when runs are equally distributed along grey levels
IBSI
GLCM_FirstMeasureOfInformationCorrelation
The IMC1 is related to the entropy of the images and gives information on how a pixel value is correlated to its neighbourod
IBSI
AreaDensityAEE
The area density approximate enclosing elipsoide is defined as the ratio between the area and the surface of the elipsoide having as axes the 3 eigenvectors from PCA. The surface of the elipsoide is approximated using infinite series
NeighbourhoodGreyLevelDependenceMatrix
The neighbouring grey level dependence matrix (NGLDM) captures the coarseness of the overall texture and is rotationally invariant. Defined in https://doi.org/10.1016/0734-189X(83)90032-4
IBSI
IntensityHistogramMinimumGradientIntensity
Minimum intensity along the gradient
IBSI
Volume
The volume V is calculated from the ROI mesh as indicated in https://doi.org/10.1109/ICIP.2001.958278
IBSI
GLDZM_SmallDistanceLowGreyLevelEmphasis
This feature emphasises runs in the upper left quadrant of the GLDZM, where small zone distances and low grey levels are located
HLL
IBSI
GLDZM_LowGreyLevelZoneEmphasis
Instead of small zone distances, low grey levels are emphasised
LanczosInterpolation
This filter uses a convolution kernel to interpolate the pixels of the input image. The kernel is based on the sampling function (sinc)
HHL
IBSI
InterquartileRange
The interquartile range (IQR) is defined as P75 - P25, where P75 is the 75th percentile of the image matrix, and P25 is the 25th
IBSI
Compactness1
The Compactness defines the deviation of the ROI volume from a perfect sphere. Compactness 1 is a measure of how compact, or sphere-like or sphere like the volume is
IBSI
IntensityHistogramQuartileCoefficientOfDispersion
The ratio of the difference between the 75 and 25 percentile and their sum
PixelOperations
This class includes all the filters which apply basic operations on one / multiple pixels of an image
IBSI
NGLDM_LowDependenceEmphasis
This feature emphasises low neighbouring grey level dependence counts
V10minusV90
Difference between the volume at intensity fraction 10 and volume at intensity fraction 90
Lloyd-Max
The Lloyd-Max algorithm is an iterative clustering method that seeks to minimise mean squared discretisation errors (Max, 1960; Lloyd, 1982).
2DMerging
Merged directions per slice and then averaged
ZeroOrder
Features which are directly derived from imaging properties like for example the SUV (Standard Uptake Volume) for PET
SUVMean
SUVMax
IBSI
VolumeDensityMVEE
The volume density minimum volume enclosing ellipsoid is the ratio between the volume and the minimum volume enclosing ellipsoid, calculated as in https://doi.org/10.1016/j.dam.2007.02.013
IBSI
GLRLM_RunLengthVariance
This feature estimates the variance in runs for run lengths
IBSI
MajorAxisLength
The major axis length is defined as twice the semi-axis length, dtermined using the largest eigenvalue obtained by principal component analysis (PCA) on the point set of voxel centers
DifferenceAverage
IBSI
GLCM_DifferentAverage
IBSI
NGLDM_LowGreyLevelCountEmphasis
Instead of low neighbouring grey level dependence counts, low grey levels are emphasised
IBSI
IntensityHistogramEntropy
It is the Shannon entropy of the histogram
GreyLevelRunLengthMatrix
The Grey level run length matrix (GLRLM) like the grey level co-occurrence matrix, GLRLM also assesses the distribution of discretised grey levels in an image or in a stack of images. However, instead of assessing the combination of levels between neighbouring pixels or voxels, GLRLM assesses grey level run lengths. Run length counts the frequency of consecutive voxels with discretised grey level i along direction delta
IBSI
GLCM_JointEntropy
As defined in http://dx.doi.org/10.1109/TSMC.1973.4309314
AffineTransformations
This class includes all the filters which apply affine transformations (e.g. rotations) which can be applied to an image
HLH
Statistical
The statistical features describe how grey levels within the region of interest (ROI) are distributed
ButterworthFilter
This filter applies a Butterworth digital algorithm to an image. It can be both a low-pass or high-pass filter. The cut-off frequency needs to be specified
Inverse
Defined as 1/x
IBSI
GLCM_JointVariance
Also called sum of squares as in http://dx.doi.org/10.1109/TSMC.1973.4309314
Pyradiomics
IntensityHistogramTotalEnergy
Total Energy is the value of Energy feature scaled by the volume of the voxel in cubic mm
ReSegmentationRange
Re-segmentation may be performed to remove voxels from the intensity mask that fall outside of a specified range. An example is the exclusion of voxels with Hounsfield Units indicating air and bone tissue in the tumour ROI within CT images, or low activity areas in PET images. This class requires to be specified the min and the max of the range with corresponding units
I10
The minimum intensity presents in at most 10% of the volume
NoDistanceWeighting
No weigths are applied to the computed distances
IBSI
GLDZM_ZoneDistanceNonUniformity
This features assesses the distribution of zone counts over the different zone distances. The feature value is low when zone counts are equally distributed along zone distances
MinimumBound
IBSI
GLRLM_LowGreyLevelRunEmphasis
The LGLRE measures the distribution of low gray level values. The LGRE is high for the image with low gray level values
TopologicalOperations
This class includes all the filters which are used to resize an image. They can be used both to increase or reduce the dimensions
IBSI
GLSZM_ZoneSizeNonUniformityNormalised
This is a normalised version of the zone size non-uniformity feature
IBSI
GLDZM_GreyLevelNonUniformity
This feature assesses the distribution of zone counts over the grey values. The feature value is low when zone counts are equally distributed along grey levels
IBSI
GLDZM_ZonePercentage
This feature assesses the fraction of the number of realised zones and the maximum num- ber of potential zones. Strongly linear or highly uniform ROI volumes produce a low zone percentage.
IBSI
LocalIntensityPeak
Local Intensity peak is defined as the mean grey level in a 1 cm3
spherical volume, centered on the voxel with the maximum grey level in the ROI intensity mask. See also https://doi.org/10.2967/jnumed.108.057307
GreyLevelDistanceZoneMatrix
The grey level distance zone matrix (GLDZM) counts the number of groups of connected voxels with a specific discretised grey level value and distance to ROI edge (Thibault et al., 2014). The matrix captures the relation between location and grey level.
IBSI
IntensityHistogramMinimumGradient
The minimum gradient of the grey level histogram
InterpolationParameters
Interpolation algoriths determine the grey level values in the interpolation grid after interpolation of the original grid. Interpolation is commonly used for texture features, to isotropic voxel sizes to be rotationally invariant
DependenceCoarseness
IBSI
GLRLM_LongRunEmphasis
The LRE measures the distribution of long runs
IBSI
GLRLM_RunLengthNonUniformity
This features assesses the distribution of runs over the run lengths (Galloway, 1975). The feature value is low when runs are equally distributed along run lengths.
I90
The minimum intensity presents in at most 90% of the volume
TopHatFilter
This filter performs morphological top-hat filtering. It computes the morphological opening and then substracts the results from the original image
IBSI
GLCM_InverseDifferenceMoment
Same as the inverse difference feature, but with lower weigths for elements that are further from the diagonal
Chebyschev
IBSI
IntensityHistogramMedian
Median value of the histogram
Dilation
The filter is used to connect features into an image. The dilation operator takes two pieces of data as input: the image to be dilated; a set of coordinate points known as structuring element
IBSI
IntensityHistogramInterquartileRange
Difference between the 75 and 25 interquartiles of the histogram
IBSI
IntensityHistogramRobustMeanAbsoluteDeviation
The intensity histogram robust mean absolute deviation is the mean restricted to grey values closer to the center of the distribution
HermiteInterpolation
This filter uses an interpolant based not only on equation for the function values, but also for the derivatives
BasisFunction
Function use to decompose an image
Linear
Linear interpolation makes use of first order polynomial
IBSI
IntensityHistogramMean
The mean gray level
MaximumBound
VoxelDimensionZ
IBSI
GLSZM_LowGreyLevelZoneEmphasis
Measures the distribution of lower gray-level size zones, with a higher value indicating a greater proportion of lower gray-level values and size zones in the image
NearestNeighbourInterpolation
This filter is a method for multivariate interpolation in one or more dimensions. The algorithm selects the value of the nearest point and does not consider the values of neighbouring points, yileding a piece-wise constant interpolant
IBSI
Median
The median intensity value
IBSI
GLSZM_GreyLevelNonUniformityNormalised
This is a normalised version of the grey level non-uniformity feature
IBSI
ApproximateVolume
In clinical practice, volumes are commonly determined by counting voxels. For volumes consisting of a large number of voxels (1000s), the differences between approximate volume and mesh-based volume are usually negligible. However for volumes with a low number of voxels (10s to 100s), approximate volume will overestimate volume compared to mesh-based volume. It is therefore only used as a reference feature, and not in the calculation of other morphological features.
IBSI
GLSZM_LargeZoneLowGreyLevelEmphasis
This feature emphasises runs in the upper right quadrant of the GLSZM, where large zone sizes and low grey levels are located
MithcellInterpolation
This filter uses a kernel to interpolate the pixels of the input image. The kernel is defined as:
1/6 [ ((12-9B-6C)|x|3 + ((-18+12B+6C)|x|2 + (6-2B)) ]; if |x| < 1;
1/6 [ ((-B-6C)|x|3 + (6B+30C)|x|2 + (-12B-48C)|x|2 + (8B+24C) ]; if 1 ≤ |x| < 2;
0 otherwise
TanH
Hyperbolic Tangent
BiLinearInterpolation
This filter uses the same concepts of Nearest Neighbour scaling excepts with interpolation. Instead of copying the neighbouring pixels (which often results in jaggy image), interpolation technique based on surrounding pixels is used to produce much smoother scaling
VoxelDimensionY
HounsfieldUnit
The Hounsfield unit (HU) scale is a linear transformation of the original linear attenuation coefficient measurement into one in which the radiodensity of distilled water at standard pressure and temperature (STP) is defined as zero Hounsfield units (HU), while the radiodensity of air at STP is defined as -1000 HU
MathOperations
This class includes all the filters which apply math operations between two / more different images
IBSI
NGLDM_DependenceCountNonUniformity
This features assesses the distribution of neighbouring grey level dependence counts over the different dependence counts. The feature value is low when dependence counts are equally distributed
IBSI
GLCM_SumAverage
The average for the cross-diagonal probabilities
IBSI
VolumeDensityOMBB
The volume density oriented mimum bounding box is the ratio between the volume and the volume of the oriented mimum bounding box. . The oriented minimum bounding box of the ROI mesh vertex
set encloses the vertex set and has the smallest possible volume
Textural
The textural features describe patterns or the spatial distribution of voxel intensities
HistoEqualization
This filter applies histogram equalization of an image. The normalized histo of the image is interpreted as the probability density function of the intensity of the image. The filter maps the histo of the input image to a new maximally-flat histo
IntensityHistogram
These features are based on the intensity histogram, whch is generated by discretising the original set of grey levels into grey level bins
2.5DMerging
merged over all slices
IBSI
GlobalIntensityPeak
The global intensity peak is the highest peak value from the mean intensity calculated with a neighbourhood for every voxel in the ROI intensity mask
IBSI
GLDZM_SmallDistanceEmphasis
This feature emphasises small distances
GLRLMSpecificParameters
V90
The maximum percentage volume with at least (90% of the max intensity)
ThresholdingBased
This class includes all the the method for segmentation which are based to appying a certain threshold to the pixel of an image. Threshold can be determined manually or looking at the gray level histo distribution
VersionStatus
PTVn(ROI)
CosH
Hyperbolic Cosine
IBSI
Minimum
The minimum intensity value
Fraction
Part of a quantitative concept
IBSI
NGLDM_GreyLevelVariance
This feature estimates the variance in dependence counts for the grey levels.
IBSI
IntensityHistogramMinimum
Minimum gray level bin
IBSI
GLCM_InverseDifferenceMomentNormalised
Normalized version of the inverse difference moment, as suggested by https://doi.org/10.5589/m02-004
VolumetricSUV
ArcTan
Inverse Tangent
IBSI
NGTDM_Strength
Feature defined in http://dx.doi.org/10.1371/journal.pone.0093600
AggregationParameters
Specifies how texture matrix are computed and aggregated
IBSI
IntensityHistogramMaximumGradient
Same feature as defined in https://doi.org/10.1016/j.radonc.2016.07.007
GTVn(ROI)
CTVn(ROI)
Equalisation
Coif1
CubicSpline
Cubic spline interpolation draw upon a larger neighbourhood to evaluate a smooth, continuous third-order polynomial at the points of the output grid.
Symmetrical
Symmetrical texture matrixes
IBSI
IntensityRange
The range of grey levels is the difference between the maximum and the minimum
IBSI
SurfaceToVolumeRatio
The surface to volume ratio is the ratio between the Surface and the Volume
MathematicalFunction
In mathematics, a function is a relation between a set of inputs and a set of permissible outputs with the property that each input is related to exactly one output.
IBSI
GLRLM_ShortRunEmphasis
The SRE measures the distribution of short runs. The SRE is highly dependent on the occurrence of short runs and it gives high value for fine texture the value of SRE is high
IBSI
Asphericity
Asphericity describes how much the ROI deviates from a perfect sphere
Pyradiomics
GLCM_Homogeneity1
Deprecated: Same as Inverse Difference
Pyradiomics
GLCM_Homogeneity2
This feature is depracated. Same as Inverse Difference Moment
Tangent
Tangent trigonometric Function. Ratio between sine and cosine
Direction
Specify direction for example for a filter
2D
Averaged over slices
IBSI
GLRLM_LongRunHighGreyLevelEmphasis
This feature emphasises runs in the lower right quadrant of the GLRLM, where long run lengths and high grey levels are located
IBSI
IntensityHistogramMaximum
Highest discretized grey level in the histogram distribution
IBSI
GLSZM_ZoneSizeVariance
This feature estimates the variance in zone counts for the different zone sizes
FunctionDistanceWeighting
Weigthening the distance by using a certain function (e.g. exponential)
RietsFilter
Max
WaveletFilter
This filter decomposes an image in wavelets. A wavelet series is a representation of a square-integrable (real- or complex-valued) function by a certain orthonormal series generated by a wavelet
IBSI
MeanAbsoluteDeviation
The mean of the absolute deviations of all voxel intensities around the mean intensitity value
IBSI
GLSZM_ZoneSizeNonUniformity
This features assesses the distribution of zone counts over the different zone sizes. The feature value is low when zone counts are equally distributed along zone sizes
IBSI
IntensityHistogramMedianAbsoluteDeviation
Histogram median absolute deviation is conceptually similar to histogram mean absolute deviation, but measures dispersion from the median instead of mean
Pyradiomics
Maximum2DDiameterSlice
Maximum 2D diameter (Slice) is defined as the largest pairwise Euclidean distance between tumor surface voxels in the row-column (generally the axial) plane.
GLCMSpecificParameters
Euclidean
Erosion
This filter is used to disconnect features and remove small ones. The erosion operator takes two pieces of data as inputs: the image to be eroded; a set of coordinate points known as structure element
IBSI
GLSZM_LargeZoneEmphasis
This feature emphasises large zones
Pyradiomics
Maximum2DDiameterColumn
Maximum 2D diameter (Column) is defined as the largest pairwise Euclidean distance between tumor surface voxels in the row-slice (usually the coronal) plane.
Min
IBSI
GLSZM_HighGreyLevelZoneEmphasis
Measures the distribution of the higher gray-level values, with a higher value indicating a greater proportion of higher gray-level values and size zones in the image
SquareCubicRatio
IBSI
GLSZM_SmallZoneEmphasis
This feature emphasises small zones
IBSI
GLSZM_SmallZoneLowGreyLevelEmphasis
This feature emphasises runs in the upper left quadrant of the GLSZM, where small zone sizes and low grey levels are located
IBSI
NGLDM_GreyLevelNonUniformityNormalised
This is the normalised version of the grey level non-uniformity feature
NoiseReduction
IBSI
IntensityHistogramRange
Difference between maximum and minimum of the histogram
IBSI
RootMeanSquare
The square root of the arithmetic mean of the squares of the values
IBSI
VolumeDensityAEE
The volume density approximate enclosing elipsoide is defined as the ratio between the volume and the volume of the elipsoide having as principal axes the eigenvenctors from the principal component analysis of the ROI
LLH
Subtraction
This filter is mainly used with images presenting similarities between them. The goal is to enhance the differences between two images. Subtraction is obtained setting as new pixel value the difference between the corresponding pixels of the two images
Decimal
GLDZMSpecificParameters
IBSI
NGLDM_DependenceCountPercentage
This feature assesses the fraction of the number of realised neighbourhoods and the max- imum number of potential neighbourhoods. The feature may be omitted as it evaluates to 1 when complete neighbourhoods are not required, which is the case under our definition
Rotation
This filter rotates an image according to a defined rotation matrix
IBSI
GLCM_JointAverage
The grey level weigthed sum of joint probabilities
2.5D
Merged over all slices
CubicMillimeter
A volume unit which is equal to 10^[-9] m^[3]
IBSI
NGTDM_Contrast
Contrast depends on the dynamic range of the grey levels as well as the spatial frequency of intensity changes
OutlierRemoval
ROI voxels with outlier intensities may be removed from the intensity mask. One method for defining outliers was suggested by Vallie`res et al. (2015) after Collewet et al. (2004).
IBSI
GLDZM_GreyLevelNonUniformityNormalised
This is the normalised version of the grey level non-uniformity feature
IBSI
LeastAxisLength
The least axis is the the axis along which the object is least extended. The least axis is twice the semi-axis length, determined using the smallest eigenvalue obtained by PCA on the point set of voxel centers
IBSI
GLCM_ClusterShade
The cluster shade is a measure of the skewness of the matrix and it is believed to gauge the perceptual concepts of uniformity. When the cluster age is high the image is asymmetric
Eulearian
IBSI
AreaDensityCH
The area density convex hull is defined as the ratio between the area and the surface of the convex hull obtained by summing the sum of the area of the faces in the convex hull
IBSI
CoefficientOfVariation
Coefficient of variation is the ratio between the standard deviation and the mean value
IBSI
GLDZM_ZoneDistanceVariance
This feature estimates the variance in zone counts for the different zone distances
RadiomicsFeature
Radiomics features are particular features which can be extracted from objects within an image (e.g. from a region of interest) and that can potentially present a prognostic value (e.g related to the survival probability of a patient).
LHL
IBSI
GLRLM_GreyLevelVariance
This feature estimates the variance in runs for the grey levels
IBSI
Percentile90
The statistical 90th percentile of the grey level distribution
IBSI
IntensityHistogramMaximumGradientIntensity
The discretized level corresponding to the maximum histogram gradient
DiscretizationParameters
Grey level discretisation or quantisation of the ROI is often required to make calculation of texture features tractable (Yip and Aerts, 2016)
IBSI
GLCM_Dissimilarity
The dissimilarity feature describes the variation of grey levels pair in an image. Dissimilarity always ranges from 0 and 1. It has maximum values when the grey level of the reference and the neighbor pixel is at the extremes of the possibile grey levels in the the texture sample
V10
The maximum percentage volume with at least (10% of the max intensity)
IBSI
GLCM_AngularSecondMoment
It represents the energy of the probability matrix.
Please note that this feature is also called Energy or Uniformity.
IBSI
Maximum
The maximum intensity value
IBSI
IntensityHistogramSkewness
The skewness measures the degree of histogram of gray levels asymmetry around the central value
IBSI
AreaUnderIVHCurve
Feature as defined in https://doi.org/10.1007/s00259-011-1845-6
Asymmetrical
Asymmetrical texture matrixes
IBSI
NGLDM_HighDependenceHighGreyLevelEmphasis
The high dependence high grey level emphasis feature emphasises neighbouring grey level dependence counts in the lower right quadrant of the NGLDM, where high dependence counts and high grey levels are located
IBSI
QuartileCoefficientOfDispersion
The quartile coefficient of dispersion is defined as the ratio between the difference of the 75 and 25 percentile divided by the sume of 75 and 25 percentile
GreyLevelSizeZoneMatrix
The grey level size zone matrix (GLSZM) counts the number of groups of connected voxels with a specific discretised grey level value and size as in https://doi.org/10.1109/TBME.2013.2284600. Voxels are con- nected if the neighbouring voxel has the same discretised grey level value
Pyradiomics
Maximum2DDiameterRow
Maximum 2D diameter (Row) is defined as the largest pairwise Euclidean distance between tumor surface voxels in the column-slice (usually the sagittal) plane.
LocalIntensity
Local intensity features consider voxel intensities within a defined neighbourhood around a center voxel
IntensityFraction
The intensity fraction "I" for grey level "i" in the range G is:
I = (i-min(G)) / (max(G) - min(G))
AbsoluteFiltering
This filter substitutes each original pixel value with its absolute value
ROImask_PartialVolumeCutOff
IBSI
GLRLM_RunPercentage
This feature assesses the fraction of the number of realised runs and the maximum number of potential runs (Galloway, 1975). Strongly linear or highly uniform ROI volumes produce a low run percentage
VolumeFraction
Percentage of total volume of ROI
IBSI
IntegratedIntensity
Integrated intensity is the average grey level multiplied by the volume. In the context of
18F-FDG-PET, this feature is called total legion glycolysis. See also https://doi.org/10.1016/j.radonc.2011.10.014
CubicMeter
CubicMilliMeter
HU
Live
Mantained
Matlab
Meter
MilliMeter
OpenSource
Proprietary
Python
SquareCentimeter
SquareMeter
SquareMilliMeter
Alpha
Beta
C
C++
CentiMeter
CubicCentiMeter
CubicMillimeterHU
VolumetricSUV
SquareCubicRatio
SquareHU
SUV
VolumetricHU
SquareSUV