Neural network-assisted analysis and microscopic rescreening in presumed negative cervical cytologic smears - A comparison

OBJECTIVE: To compare cytologists' defection of abnormalities when using neural network-assisted (NNA) review, as employed by the PAPNET Testing System and to compare the effectiveness of this mode of review to that of unassisted, conventional rescreening of cervical smears initially diagnosed as negative. STUDY DESIGN: The study was undertaken as part of a multicenter clinical trial involving over 10,000 smears from 10 investigation sites (9 academic institutions and 1 private laboratory). Using a subset of "negative" control smears from three university laboratories, the false negative detection yields of NNA review (performed using the PAPNET System) and conventional microscopic rescreening (performed as part of routine quality control practice) were compared. The false negative detection yield was defined Its the percentage of rescreened negatives reclassified as abnormal. RESULTS: The results demonstrate that using NNA review, the detection yield of false negative smears, as a proportion of negative smears reexamined, is statistically significantly greater than that obtained using conventional quality control rescreening. The false negative yield generated using NNA analysis was 6.2% (142/2293) versus 0.6% (82/13761) for conventional rescreening. A statistically significant improvement in identification of abnormality is observed for NNA review as opposed to unassisted rescreening despite constraining the comparison in the following ways: (1) comparing the yields of rescreening of negative smears obtained from the same time intervals for both methods, (2) comparing the yields of rescreening of negative smears obtained from the years after the Clinical Laboratory Improvement Act (1990 and 1991) for both methods, and (3) disregarding the identification of atypical squamous cells of undetermined significance/atypical glandular cells of undetermined significance cases and comparing only the identification of squamous intraepithelial lesions using the two methods. CONCLUSION: Using neural network-assisted review, cytologists uncovered a significantly higher proportion of previously undetected cervical abnormalities per smear reexamined than they did using unassisted, conventional rescreening.