Assessment of the Incremental Benefit of Computer-Aided Detection (CAD) for Interpretation of CT Colonography by Experienced and Inexperienced Readers

Abstract

ObjectivesTo quantify the incremental benefit of computer-assisted-detection (CAD) for polyps, for inexperienced readers versus experienced readers of CT colonography.Methods10 inexperienced and 16 experienced radiologists interpreted 102 colonography studies unassisted and with CAD utilised in a concurrent paradigm. They indicated any polyps detected on a study sheet. Readers’ interpretations were compared against a ground-truth reference standard: 46 studies were normal and 56 had at least one polyp (132 polyps in total). The primary study outcome was the difference in CAD net benefit (a combination of change in sensitivity and change in specificity with CAD, weighted towards sensitivity) for detection of patients with polyps.ResultsInexperienced readers’ per-patient sensitivity rose from 39.1% to 53.2% with CAD and specificity fell from 94.1% to 88.0%, both statistically significant. Experienced readers’ sensitivity rose from 57.5% to 62.1% and specificity fell from 91.0% to 88.3%, both non-significant. Net benefit with CAD assistance was significant for inexperienced readers but not for experienced readers: 11.2% (95%CI 3.1% to 18.9%) versus 3.2% (95%CI -1.9% to 8.3%) respectively.ConclusionsConcurrent CAD resulted in a significant net benefit when used by inexperienced readers to identify patients with polyps by CT colonography. The net benefit was nearly four times the magnitude of that observed for experienced readers. Experienced readers did not benefit significantly from concurrent CAD.

Tracking Eye Gaze during Interpretation of Endoluminal Three-dimensional CT Colonography: Visual Perception of Experienced and Inexperienced Readers

“…Because 97% of our observations (760 of 787) included a close pursuit of polyps, scanning errors can only account for few missed polyps in our study. The rate of scanning errors was similar for experienced and inexperienced readers, and it cannot account for the greater accuracy of experienced readers in this study and others (10)(11)(12). We found that readers almost always examined polyps by a series of multiple pursuits, which suggested that readers recognize a lesion as visually important in most observations, and the errors are because of decision.…”

“…and E.H.) from time consuming compared with standard abdominopelvic CT (7), and it requires considerable training (9). Experienced readers are more accurate than less experienced readers (10)(11)(12), but understanding of why they are more accurate is limited. Knowledge of the deficiencies in interpretation by inexperienced readers could improve the focus and structure of training.…”

Interobserver Variation of Colonic Polyp Measurement at Computed Tomography Colonography

“…Variable outcomes have been demonstrated associating CTC CAD measurements, reader experience, and observer sensitivity. Authors have reported an increase in reader sensitivity with CAD, particularly in those who are inexperienced [30–33] or moderately experienced in CTC for polyp detection, both as concurrent and second reader [34], while experienced readers in comparison did not benefit significantly from concurrent CAD [35]. Due to these inconsistencies, a second read CAD is generally recommended [27].…”

CT colonography: implementation, indications, and technical performance – a follow-up national survey

“…CAD is a software developed to mark suspicious findings and thereby improve the diagnostic performance of the reader. CAD systems are primarily designed for polyp detection, and have been shown to improve detection of polyps, at least for less experienced readers, with a limited decrease in specificity (10,11). With increasing experience in reading CTC, however, CAD has a smaller impact on the final evaluation (10,11).…”