| aucMRMCcluster {mitoticFigureCounts} | R Documentation |
The results of a multi-reader multi-case (MRMC) analysis of the auc for each scanner.
The MRMC analysis is accomplished by the OR method (Obuchowski and Rockette:
Obuchowski1995_Commun-Stat-Simulat_v24p285).
The function used is mrmcAnalysisOR.
Since the data is binary,
auc is the average of sensitivity and specificity or half of (Youden's index + 1).
Sensitivity is defined as the number of MFs
detected by an observer divided by the number of true MFs.
Specificity is defined as one minus the false-positive fraction,
where the false-positive fraction is the number of false MFs that were positively marked,
divided by the total number of false MFs.
Furthermore, we account for the fact that there are multiple observations
per case (multiple ROIs per WSI, clustered data: Obuchowski1997_Biometrics_v53p567)
when calculating the reader by modality covariances that are used in the OR method.
The function used is doAUCcluster
aucMRMCcluster
An object of class list of length 4.
The results of this script are a list of MRMC AUC analysis results for 4 scanners.
Scanner.A,
Scanner.B,
Scanner.C,
Scanner.D
Each scanner results is itself a list contains the MRMC analysis results for the null hypothesis that the scanner accuracy is the same as the microscope accuracy:
F_stat, F-statistic
df.H, Degrees of freedom
p, p-value
theta.i, [2] Accuracy of the microscope and scanner
df.sgl, [2] Degrees of freedom for the microscope and the scanner
se.i, [2] Standard error of the microscope and scanner accuracy
se.dif, Standard erro of the microscope and scanner accuracy difference
covOR, [12] Obuchowski and Rockette covariances:
varR, varTR, cov1, cov2, cov3, varE, varR.1, varR.2, cov2.1, cov2.2, varE.1, varE.2
theta.hat, [2x5] Accuracy of each reader for the microscope (row 1)
and accuracy of each reader for the scanner (row 2)
cov.hat, [10x10] Reader by modality (microscope and scanner) covariance matrix
botCI, [2] Bottom 95
topCI, [2] Top 95
#' The analysis results are used to produce:
Table 4: Accuracy for all readers and observation methods
Figure 3: Accuracy (average of sensitivity and specificity) for each viewing mode averaged over all the readers with 95 The asterisks indicate that the difference in accuracy of the viewing mode compared to that of microscopy is statistically significant. All analyses account for the correlations and variability from the readers reading the same ROIs.