viperObservations

Description

The VIPER data sets were created by aggregating all the raw reader data files from the FDA VIPER study: Validation of Imaging Premarket Evaluation and Regulation (VIPER). The data sets differ in the truth labels. See the documentation details.

Usage

viperObs

viperObs365

viperObs455

Format

An object of class data.frame with 20382 rows and 35 columns.

Details

viperObs is a simplified and merged version of viperObs455 (truth labels based on cancer at 455 days) and viperObs365 (truth labels based on cancer at 365 days).

The images used in this study were selected from the DMIST images [Pisano2005_NEJM_v353p1773]. The VIPER study concept 2011. Data collection September 2013 to August 2015. Please refer to journal paper for details.

The original data frames had the following 35 variables:

• CaseReadOrder
  

• readerID.relative

• Study
  

• Track
  

• ReadSetLetter
  

• Ctype
  

• Compare.CaseIDs
  

• Initial.Comparison
  

• caseID
  

• redcap_event_name
  

• readerID
  

• form_start_datetime
  

• modality
  

• study
  

• viewbox
  

• num_img
  

• recall
  

• miss_img
  

• score_start_datetime
  

• abnormality_type
  

• recall_scale
  

• score_end_datetime
  

• Group
  

• norecall_scale
  

• set_letter
  

• track
  

• quality
  

• UID
  

• score
  

• caseGroup

• desc
  

• read_order
  

• Comment
  

• RecallRegion
  

• count