BRISC::Core::MarkovRandom Class Reference


Static Public Member Functions
static void	PerformMarkov (LIDCNodule nodule)
	Performs Markov analysis on the given nodule and saves the results to the appropriate LIDCNodule fields.
static double[][,]	FindParameters (int[,] segData)
	Finds the MRF parameters.
Static Public Attributes
static double	histMinVal = 0
	Minimum value for histogam binning. All values below this will be binned into the bottom bin.
static double	histMaxVal = 0
	Maximum value for histogram binning. All values above this will be binned into the top bin.
static int	binSize = 512
	Number of bins in the response histograms.
Static Private Member Functions
static void	normalize (ref int[,] segData)
	Normalizes the image with zero mean.
static void	deriveNewFeatures (int[,] segData, int x, int y, int pad, ref double[][,] markovParams)
	Calculates new features from the beta parameters.
static void	calcSumCorrVect (int[,] segData, int x, int y, ref GeneralMatrix sum_corr, ref GeneralMatrix sum_vector)
	Function calculates the sum of the correlation and vectors.
static double	calcSumVar (int[,] segData, int x, int y, ref GeneralMatrix parameters)
	Sums up the variances and the first part of the beta parameter.
Static Private Attributes
static int	pad = 4
	Padding needed around the image.

Detailed Description

Member Function Documentation

static void BRISC::Core::MarkovRandom::calcSumCorrVect	(	int	segData[,],
		int	x,
		int	y,
		ref GeneralMatrix	sum_corr,
		ref GeneralMatrix	sum_vector
	)			`[inline, static, private]`

Function calculates the sum of the correlation and vectors.

Parameters:

	segData	Original image
	x	Current column index
	y	Current row index
	sum_corr	Matrix to store the sum of the correlation
	sum_vector	Matrix to store the sum of the vectors

static double BRISC::Core::MarkovRandom::calcSumVar	(	int	segData[,],
		int	x,
		int	y,
		ref GeneralMatrix	parameters
	)			`[inline, static, private]`

Sums up the variances and the first part of the beta parameter.

Parameters:

	segData	Original image
	x	Current column index
	y	Current row index
	parameters	Matrix to store beta parameters

static void BRISC::Core::MarkovRandom::deriveNewFeatures	(	int	segData[,],
		int	x,
		int	y,
		int	pad,
		ref double	markovParams[][,]
	)			`[inline, static, private]`

Calculates new features from the beta parameters.

Parameters:

	segData	Original image
	x	Current column index
	y	Current row index
	pad	Padding around the image
	markovParams	Passed in as beta values. Returned as new features

static double [][,] BRISC::Core::MarkovRandom::FindParameters ( int segData[,] ) [inline, static]

Finds the MRF parameters.

Parameters:

segData

Image to run calculations on

static void BRISC::Core::MarkovRandom::normalize ( ref int segData[,] ) [inline, static, private]

Normalizes the image with zero mean.

Parameters:

segData

Image to normalize

Results are worse if we normalize. Not sure why this is, reguardless we are not using this function. It's here just in case.

static void BRISC::Core::MarkovRandom::PerformMarkov ( LIDCNodule nodule ) [inline, static]

Performs Markov analysis on the given nodule and saves the results to the appropriate LIDCNodule fields.

Parameters:

nodule

LIDCNodule to run Markov on

Static Public Member Functions