Time-Series Analysis of Human Interpretation Data in Mammography
Abstract
Recent research has documented that the human observer is a significant source of interpretation errors in mammography in the U.S. However, it has yet to be determined whether or not the rate or likelihood of radiologist-specific error changes across the length of time the radiologist has been reading during a single session, or across the cumulative time the radiologist reads in a year. The purpose of this study was to apply basic methods from the statistical analysis of time series in order to gain novel insights into the characteristics of the human interpretation of mammograms.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2005
- Accession Number
- ADA434583
Entities
People
- Craig A. Beam
- Edward A. Sickles
- Emily F. Conant
- Harold L. Kundel
- Ji-Hyun Lee
- Patricia A. Romily
Organizations
- H. Lee Moffitt Cancer Center & Research Institute