We assume there is a quality or related measure that will let us compute distances between different time points and that these distances relate to how similar two images are. Note that the measurement may not relate to quality at all but should map images into a similar metric space based on appearance or some other objective quality that leads to the same value when images are the same and changes continuously as images differ.
assembleTestRetest(
dataIn,
qualitycolumns,
subjectID,
visitID,
whichVisit,
measureToRepeat,
uniqueID,
covariates,
nozeroes = FALSE
)Input data frame with the old name within
the name(s) of quality-related column measurements
the unique subject id column name
the column name that gives the visit / date; typically we look for data the exists on the same visit.
optional visit to use for the analysis
the measurement to assemble for computing trt stats (e.g. ICC)
optional column to contatenate to trt dataframe
optional additional column name(s) to add to dataframe
optional boolean - do not allow zero distance time
data frame with test-retest friendly organization ... trt0 and trt1 show the row indices of the test retest data
mydf = generateSubtyperData( 100 )
rbfnames = names(mydf)[grep("Random",names(mydf))]
mytrt = assembleTestRetest( mydf, rbfnames[1], 'Id', 'visit' )