applyflowchart: Annotate and label each reader mark based on a decision tree

In rickeycarter/mrmctools: Statistical Tools to Support Multi-Reader, Multi-Case (MRMC) Study Designs

Description

This function normalizes the readerworkstation file (the data that is passed to this function) and prepares it for further analysis. The coding used to designate the primary task ROIs is a requried parameter. The includeA and includeC parameters define some special predetermined exclusions from the function. See the vignette for further elaboration of these parameters.

Usage

applyflowchart(rawdata, PTcodes, PTCthreshold, includeA = T, includeC = T)

Arguments

rawdata	The output datafile from the function readworkstation file.
PTcodes	An R set object that contains the list of database ID codes that are included in the primary task. If a reader mark is in this set, the ObsPTLesion variable in the output dataset will be coded as a 1. Else, it will be coded as 0. The same logic is applied to the classification of reference lesions and the outputted variable name is RefPTLesion.
PTCthreshold	A numeric value used to determine the lower limit for any primary task confidence score to be considered in the analysis. The limit is applied as a greater than equal to. So specification of 1 will include all reader ROIs with a score greater than or equal to 1.
includeA	A parameter, defaulted to True, that allows one to exclude type A reader marks. These are reader marks where they do not match any lesions in the reference set and the reader classifies them as a non-primary task lesion. Given the reader's confidence could be non-negigble, under most circumstations, these reader marks should be retained. Setting this to false will exclude these markings from all subsequent analyses.
includeC	A parameter, defaulted to True, that allows one to exclude type C reader marks. These are reader marks where both the reader and reference localize the same non-primary task lesion. Given the reader's confidence could be non-negigble, under most circumstations, these reader marks should be retained. Setting this to false will exclude these markings from all subsequent analyses.

Value

A new dataframe that appends several helper columns onto the input dataset.

rickeycarter/mrmctools documentation built on May 6, 2019, 8:59 p.m.