CalciumScoring.Agatston — Type

Agatston

struct Agatston <: CalciumScore end

Lets Julia know (via multiple dispatch) that the algorithm of choice when calculating the calcium score should be Agatston's algorithm

CalciumScoring.CalciumScore — Type

CalciumScore

abstract type CalciumScore end

Main type for all calcium scoring algorithms in this package

CalciumScoring.SpatiallyWeighted — Type

SpatiallyWeighted

struct SpatiallyWeighted <: CalciumScore end

Lets Julia know (via multiple dispatch) that the algorithm of choice when calculating the calcium score should be the Spatially Weighted Calcium Scoring algorithm.

CalciumScoring.VolumeFraction — Type

VolumeFraction

struct VolumeFraction <: CalciumScore end

Lets Julia know (via multiple dispatch) that the algorithm of choice when calculating the calcium score should be the volume fraction algorithm

CalciumScoring._percentage_calcium — Method

_percentage_calcium

_percentage_calcium(voxel_intensity, hu_calcium, hu_heart_tissue)

Find the percentage of calcium within one voxel given the voxel intensity voxel_intensity, the Hounsfield unit associated with a known calcium density hu_calcium, and the Hounsfield unit associated with background heart tissue hu_heart_tissue

CalciumScoring._weight_thresholds — Method

_weight_thresholds

_weight_thresholds(kV, max_intensity)

This code defines a function named _weight_thresholds that calculates the weight threshold for the Agatston scoring algorithm. The function takes two input parameters: kV and maxintensity. The kV parameter represents the CT scan's kilovoltage peak (kVp) value, and `maxintensity` represents the maximum intensity of a calcified lesion in Hounsfield units (HU).

The function checks the validity of the input kV value and ensures that it is one of the accepted values: 70, 80, 100, 120, or 135. If the input value does not match any of the accepted values, an error message is displayed, and the program is halted.

The function then calculates the weight threshold based on the input kV and max_intensity. It uses a series of if-else statements to determine the weight threshold depending on the kV value and the given range of max_intensity values.

The function returns the weight threshold as an integer value between 0 and 4.

Inputs

  • kV: The kilovoltage peak (kVp) value of the CT scan (Accepted values: 70, 80, 100, 120, or 135)
  • max_intensity: The maximum intensity of a calcified lesion in Hounsfield units (HU)

Returns

  • weight: The weight threshold for the Agatston scoring algorithm (integer value between 0 and 4)

Reference

CalciumScoring.score — Method

score (VolumeFraction)

score(vol::AbstractArray, hu_calcium, hu_heart_tissue, alg::VolumeFraction)

score(vol::AbstractArray, hu_calcium, hu_heart_tissue, voxel_size, alg::VolumeFraction)

score(vol::AbstractArray, hu_calcium, hu_heart_tissue, voxel_size, density_calcium, alg::VolumeFraction)

score(vol::AbstractMatrix, hu_calcium, hu_heart_tissue, alg::VolumeFraction)

score(vol::AbstractMatrix, hu_calcium, hu_heart_tissue, voxel_size, alg::VolumeFraction)

score(vol::AbstractMatrix, hu_calcium, hu_heart_tissue, voxel_size, density_calcium, alg::VolumeFraction)

Calculate the number of voxels that are calcified, the volume of the calcification, or the mass of the calcification depending on the provided arguments.

Inputs

  • vol: An input region of interest containing all potential calcium.
  • hu_calcium: The Hounsfield unit associated with a known calcium density.
  • hu_heart_tissue: The Hounsfield unit associated with background heart tissue.
  • voxel_size: (optional) The size of an individual pixel/voxel.
  • density_calcium: (optional) The density of the previously mentioned calcium.
  • alg::VolumeFraction: The algorithm to be used for the calculation.

Returns

  • When called with vol, hu_calcium, hu_heart_tissue, and alg, this function calculates and returns the number of voxels that are calcified.
  • When called with vol, hu_calcium, hu_heart_tissue, voxel_size, and alg, this function calculates and returns the volume of the calcification.
  • When called with vol, hu_calcium, hu_heart_tissue, voxel_size, density_calcium, and alg, this function calculates and returns the mass of the calcification.

References

Coronary artery calcium mass measurement based on integrated intensity and volume fraction techniques

CalciumScoring.score — Method

score (SpatiallyWeighted)

score(vol::AbstractMatrix, calibration, alg::SpatiallyWeighted)

score(vol::AbstractMatrix, μ, σ, alg::SpatiallyWeighted)

score(vol::AbstractArray, calibration, alg::SpatiallyWeighted)

score(vol::AbstractArray, μ, σ, alg::SpatiallyWeighted)

Calculate the calcium score via the Spatially Weighted Calcium Scoring algorithm. This method avoids thresholding by weighting each voxel based on a previous calibration (usually 100 mg/cc) as outlined in the MESA study.

The function can be called with different parameters:

  1. vol::AbstractMatrix, calibration, alg::SpatiallyWeighted - Applies the algorithm to a 2D volume with a calibration array.
  2. vol::AbstractMatrix, μ, σ, alg::SpatiallyWeighted - Applies the algorithm to a 2D volume with given mean μ and standard deviation σ.
  3. vol::AbstractArray, calibration, alg::SpatiallyWeighted - Applies the algorithm to a 3D volume with a calibration array.
  4. vol::AbstractArray, μ, σ, alg::SpatiallyWeighted - Applies the algorithm to a 3D volume with given mean μ and standard deviation σ.

Inputs

  • vol: input volume containing just the region of interest. This can be a 2D (AbstractMatrix) or 3D (AbstractArray) volume.
  • calibration: a previous calibration for weighting each voxel.
  • μ, σ: given mean and standard deviation.
  • alg::SpatiallyWeighted: Spatially Weighted scoring algorithm SpatiallyWeighted().

Returns

  • sw_score: total Spatially Weighted score, providing a continuous and improved measure of atherosclerosis compared to existing coronary artery calcium scoring methods.

References

An alternative method for quantifying coronary artery calcification: the multi-ethnic study of atherosclerosis (MESA)

Spatially Weighted Coronary Artery Calcium Score and Coronary Heart Disease Events in the Multi-Ethnic Study of Atherosclerosis

CalciumScoring.score — Method

score (Agatston)

score(vol, spacing, alg::Agatston; kV=120, min_size_mm=1)
score(vol, spacing, mass_cal_factor, alg::Agatston; kV=120, min_size_mm=1)

Calculate the calcium score via the traditional Agatston scoring technique, as outlined in the original paper. Energy (kV) specific thresholds are determined based on previous publications. Also, it converts the Agatston score to a calcium mass score via the mass_cal_factor if provided.

Inputs

  • vol: input volume containing just the region of interest
  • spacing: known pixel/voxel spacing
  • alg::Agatston: Agatston scoring algorithm Agatston()
  • mass_cal_factor: (optional) a factor to calibrate the calculated mass
  • kwargs:
    • kV=120: energy of the input CT scan image
    • min_size=1: minimum connected component size (see label_components)

Returns

  • agatston_score: total Agatston score
  • volume_score: total calcium volume via Agatston scoring
  • abs_mass_score: (optional) total calcium mass after calibration, returned only when mass_cal_factor is provided

References

Quantification of coronary artery calcium using ultrafast computed tomography

Ultra-low-dose coronary artery calcium scoring using novel scoring thresholds for low tube voltage protocols—a pilot study

CalciumScoring.score — Method

score (Integrated)

score(S_Bkg, S_Obj, algorithm::Integrated)

score(S_Bkg, S_Obj, size, algorithm::Integrated)

score(S_Bkg, S_Obj, size, ρ, algorithm::Integrated)

Use the integrated intensity approach to find the total number of object voxels/volume/mass contained in vol.

The function can be called with different parameters:

  1. S_Bkg, S_Obj, algorithm - Returns the total number of object voxels.
  2. S_Bkg, S_Obj, size, algorithm - Returns the total object volume.
  3. S_Bkg, S_Obj, size, ρ, algorithm - Returns the total object mass.

Inputs

  • S_Bkg: pure background signal intensity in the vol
  • S_Obj: pure object signal intensity in the vol
  • size: (optional) size of the voxels in the given vol (only for the 2nd and 3rd variations)
  • ρ: (optional) density of the given object of interest contained within the vol (only for the 3rd variation)
  • algorithm: the Integrated algorithm which accounts for noise and partial volume effect by integrating the intensity of the entire volume vol

References

Integrated intensity-based technique for coronary artery calcium mass measurement: A phantom study

Accurate quantification of vessel cross-sectional area using CT angiography: a simulation study