© RSNA, 2008
CT Colonography and Computer-aided Detection: Effect of False-Positive Results on Reader Specificity and Reading Efficiency in a Low-Prevalence Screening Population
Table E1. Reader Confidence in Overall Study Classification as Normal or Abnormal (6-mm Polyp Size Threshold) without and with CAD
Note.—Numbers in parentheses are standard deviations, except where noted.
* Mean change in confidence expressed as CAD readings minus readings without CAD. Numbers in parentheses are 95% confidence intervals.
† P values calculated by using paired t test.
‡ Significant P value.
§ Confidence same without and with CAD for all subjects for this reader.
 |
| Figure E1: Distribution of number of CAD FP marks in studies with normal findings (n = 45). Line = polynomial trend line. |
 |
 |
| a. |
b. |
| Figure E2: (a) Two-dimensional transverse and (b) 3D endoluminal CT colonographic images show 10-mm hepatic flexure polyp (arrow) on haustral fold (arrowhead). CAD correctly depicts polyp (square). Both marks were considered CAD true-positive marks. Three readers correctly identified polyp without CAD and all four identified polyp after CAD application. |
