Please use this identifier to cite or link to this item:
|Scopus||Web of Science®||Altmetric|
|Title:||Effectiveness of autopapsystem location-guided screening in the evaluation of cervical cytology smears|
|Citation:||Diagnostic Cytopathology, 2004; 31(2):94-99|
|Mark W. Stevens, Amanda J. Milne, Ian H. Parkinson, Walter W. Nespolon, Nicola L. Fazzalari, Neeru Arora and Thomas J. Dodd|
|Abstract:||Location-guided screening is a feature of the AutoPap primary screening software. Areas of the smear most likely to contain an abnormality are identified for prompt review by the cytotechnologist, thereby facilitating diagnostic accuracy and reducing laboratory workload. A two-armed retrospective study comprising 6,000 conventional smears was undertaken to compare this approach with the current practice of full manual screening of conventional smears. Discrepant diagnoses between the two study arms were subject to an internal discrepancy review process to determine the final truth diagnosis. Analysis of the results show that AutoPap location-guided screening is at least equivalent to current practice when detecting high-grade or suspected high-grade smears. However, the device does not detect low-grade abnormalities, unsatisfactory smears, an endocervical component or organisms, as well as standard screening. The device also assigns a numerical score to each slide, with abnormal smears allocated a higher rank. Slide ranking was found to be of value in triaging abnormal smears for prompt screening and reporting. The performance of the primary screening software was found to be comparable to previous studies, with the majority of abnormal smears being selected by the instrument for manual review.|
|Keywords:||Humans; Vaginal Smears; Sensitivity and Specificity; Automation; Software; Uterine Cervical Neoplasms; Female|
|Appears in Collections:||Pathology publications|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.