RJafroc (version 1.0.1)

FitRsmRoc: Fit the radiological search model (RSM) to ROC data

Description

Fit an RSM-predicted ROC curve to a binned dataset

Usage

FitRsmRoc(dataset, lesDistr, trt = 1, rdr = 1)

Arguments

dataset

The binned dataset containing the data

lesDistr

The lesion distribution matrix

trt

The desired treatment, default is 1

rdr

The desired reader, default is 1

Value

The return value is a list with the following elements:

mu

The mean of the diseased distribution relative to the non-diseased one

lambdP

The Poisson parameter describing the distribution of latent NLs per case

nuP

The binomial success probability describing the distribution of latent LLs per diseased case

zetas

The RSM cutoffs, zetas or thresholds

AUC

The RSM fitted ROC-AUC

StdAUC

The standard deviation of AUC

NLLIni

The initial value of negative LL

NLLFin

The final value of negative LL

ChisqrFitStats

The chisquare goodness of fit results

covMat

The covariance matrix of the parameters

fittedPlot

A ggplot2 object containing the fitted operating characteristic along with the empirical operating points. Use print to display the object

Details

In the RSM: (1) The (random) number of latent NLs per case is Poisson distributed with mean parameter lambdaP, and the corresponding ratings are sampled from \(N(0,1)\). The (2) The (random) number of latent LLs per diseased case is binomial distributed with success probability nuP and trial size equal to the number of lesions in the case, and the corresponding ratings are sampled from N(\(mu\),1). (3) A latent NL or LL is actually marked if its rating exceeds the lowest threshold zeta1. To avoid clutter error bars are only shown for the lowest and uppermost operating points. MLE ROC algorithms require binned datasets. Use DfBinDataset to perform the binning prior to calling this function. Because of the extra parameter, and the requirement to have five counts, the chi-square statistic often cannot be calculated.

References

Chakraborty DP (2006) A search model and figure of merit for observer data acquired according to the free-response paradigm. Phys Med Biol 51, 3449-3462.

Chakraborty DP (2006) ROC Curves predicted by a model of visual search. Phys Med Biol 51, 3463--3482.

Chakraborty DP (2017) Observer Performance Methods for Diagnostic Imaging - Foundations, Modeling, and Applications with R-Based Examples, CRC Press, Boca Raton, FL. https://www.crcpress.com/Observer-Performance-Methods-for-Diagnostic-Imaging-Foundations-Modeling/Chakraborty/p/book/9781482214840

Examples

# NOT RUN {
## Test with included ROC data (some bins have zero counts)
lesDistr <- UtilLesionDistribution(dataset02)
retFit <- FitRsmRoc(dataset02, lesDistr)
print(retFit$fittedPlot)

## Test with included degenerate ROC data
lesDistr <- UtilLesionDistribution(datasetDegenerate)
retFit <- FitRsmRoc(datasetDegenerate, lesDistr);print(retFit$fittedPlot)

## Test with single interior point data
fp <- c(rep(1,7), rep(2, 3))
tp <- c(rep(1,5), rep(2, 5))
dataset <- Df2RJafrocDataset(fp, tp)
lesDistr <- UtilLesionDistribution(dataset)
retFit <- FitRsmRoc(dataset, lesDistr);print(retFit$fittedPlot)

## Test with two interior data points
fp <- c(rep(1,7), rep(2, 5), rep(3, 3))
tp <- c(rep(1,3), rep(2, 5), rep(3, 7))
dataset <- Df2RJafrocDataset(fp, tp)
lesDistr <- UtilLesionDistribution(dataset)
retFit <- FitRsmRoc(dataset, lesDistr);print(retFit$fittedPlot)


## Test with three interior data points
fp <- c(rep(1,12), rep(2, 5), rep(3, 3), rep(4, 5)) #25
tp <- c(rep(1,3), rep(2, 5), rep(3, 7), rep(4, 10)) #25
dataset <- Df2RJafrocDataset(fp, tp)
lesDistr <- UtilLesionDistribution(dataset)
retFit <- FitRsmRoc(dataset, lesDistr);print(retFit$fittedPlot)


# }