The lesionID distribution of a dataset or a supplied 1D-array

The lesionID distribution of a dataset or of a supplied 1D-array. The lesionID field is described in the format of the Excel input file, see QuickStart (use Command click to view the link).

UtilLesDistr(dsOrArr)

Arguments

dsOrArr

A dataset object from which the lesion distribution is extracted or a 1D-array specifying the lesion distribution.

Value

A data frame containing the number of lesionIDs per case, lesID, and the frequency distribution of the lesionIDs, Freq.

Details

Apart from ratings two characteristics of an FROC dataset affect the FOM: the distribution of lesionIDs per case and the lesion weights. This function addresses the former. The latter is addressed in UtilLesWghts. The return value is a dataframe containing 2 equal length vectors: the lesionID labels and the corresponding fractions of lesions per diseased case in the dataset. For ROC or LROC data this vector is always c(1), since all diseased cases contain one lesion. For FROC data the first element of the dataframe, i.e., lesID, contains the numbers of lesions per diseased case. The second element, i.e., Freq, contains the fraction of dis. cases containing one lesion, the fraction containing two lesions, etc. See PlotRsmOpChrs for a function that depends on LD.

Examples

res <- UtilLesDistr(dataset01) # FROC dataset
##    lesID       Freq
## 1      1 0.93258427
## 2      2 0.06741573
## In the Excel input file, 93 percent of lesions have lesionID = 1
## and the rest have lesionID = 2


res <- UtilLesDistr(dataset02) # ROC dataset
##       lesID Freq
## 1         1    1
## In the Excel input file all dis. cases have one lesion

res <- UtilLesDistr(datasetCadLroc) # LROC dataset
##       lesID Freq
## 1         1    1
## In the Excel input file all dis. cases have one lesion

res <- UtilLesDistr(c(0.5, 0.3, 0.1, 0.1))
##      lesID Freq
## 1        1  0.5
## 2        2  0.3
## 3        3  0.1
## 4        4  0.1
## An example of array input; 50 percent of the cases have lesionID = 1,
## 30 percent have lesionID = 2, etc.

res <- UtilLesDistr(dataset11) ## big froc dataset
##     lesID        Freq
## 1       1 0.217391304
## 2       2 0.200000000
## 3       3 0.113043478
## 4       4 0.086956522
## 5       5 0.043478261
## 6       6 0.095652174
## 7       7 0.052173913
## 8       8 0.069565217
## 9       9 0.017391304
## 10     10 0.026086957
## 11     11 0.026086957
## 12     12 0.026086957
## 16     16 0.017391304
## 20     20 0.008695652

## TBA!! This dataset has lots of lesions (3D imaging for lung cancer).
## The lesionIDs range from 1 to 20 with a few missing,
## e.g., lesionID = 13 is not present in any diseased case.
## Cases with lesionID = 1 have frequency 0.217, those with lesionID = 16
## have frequency 0.174, those with lesionID = 20 have frequency 0.00870, etc.