PAPNET-assisted primary screening of conventional cervical smears.

The PAPNET System is the only device with a neural-network-based-artificial intelligence to detect and show the images of abnormal cells on the monitor to be evaluated in an interactive way. We effectively used the PAPNET in rescreening of conventional cervical smears and we detected its advantages and its disadvantages. In this paper, we report our results from PAPNET-assisted primary screening performed on 20,154 conventional smears. The smears were classified as Negative or as Review. The Negative cases were rapidly rescreened mainly near the coverslip edges, which are the slide areas not analyzed by automated devices because of focusing problems. The Review cases were fully reanalyzed by the optic microscope. In summary, 140 positive smears were detected: 57 cases showed changes due to HPV, 63 LSIL, 15 HSIL, and 5 carcinomas. Therefore, the PAPNET System was confirmed as useful in primary screening of conventional cervical samples as well as rescreening.