作者: Y. T. Van Der Schouw , H. Doornewaard , A. B. Bos , J. G. Van Den Tweel , J. D. F. Habbema
DOI:
关键词:
摘要: OBJECTIVES To assess computer-assisted (neural network based) cervical smear screening as a primary tool for the early detection of dysplasia. DESIGN Longitudinal cohort study. SETTING Cytology laboratory reviewing smears taken by general practitioners in mass program Netherlands. SUBJECTS 846 women who developed (pre-)neoplasia cervix seven years after baseline smear, and 5217 controls. INTERVENTIONS Cervical were evaluated both conventional light microscopy with use PAPNET Testing System same cytotechnologists. MAIN OUTCOME MEASURES Seven year histological cytological follow-up results obtained all from nation-wide pathology database. RESULTS Conventional diagnosed dysplasia or carcinoma 458 (54.1%) during follow-up, whereas analysis detected such lesions 462 (54.6%) these women. In control population (86.0%) women, whom revealed no dysplasia, gave false positive 210 (4.0%) 207 smears. The areas under receiver operation curves (AUC) 80% (95% confidence interval, 78 to 82%) 79% 77 81%) PAPNET-assisted screening, respectively. CONCLUSIONS has similar diagnostic value Pap when used screening.