Many of the investigations that are now carried out during imaging systems' evaluations and comparisons include interpretation (reading) studies of normal and abnormal cases that are evaluated by receiver operating characteristics (ROC) or derivatives thereof. Unfortunately, if one takes into account case and reader representativeness, most studies would require prohibitively large sample sizes and a large number of readers in order to produce adequate statistical power for a statistically significant discrimination between reading nodes in a manner that is both clinically relevant and generalizable. Approaches to case and observer selection protocols for these studies in this field have been largely arbitrary (and frequently suboptimal). The applicants propose to investigate the possibility that the implementation of specific case and reader selection protocols could allow one to focus on limited clinical questions that can be conclusively answered with an observer performance study while requiring a small effort; hence, making such studies much more feasible and reasonable to perform. During the prior effort, the applicants reported focusing their investigation on case selection issues. While they proposed to continue to investigate several aspects in this regard, the current proposal would focus on the possible effects of observer (reader) selection and training on study design and results.
Showing the most recent 10 out of 16 publications