The 2014 Bethesda System diagnostic criteria for atypical glandular cells (AGC) aid in the classification of atypical cells in cervical cytology. Anyway, AGC diagnosis remains challenging, due to low frequencies of this finding (approximately 0.5%-1% of Pap test results), abundance of AGC mimics, and significant interobserver variability. We developed an algorithm based on nuclear areas parameter that can help to differentiate AGC from Normal and Reactive glandular cells. Nuclear areas and perimeters were measured on 16 Pap smears with AGC and 18 with Reactive glandular cells of women aged between 30 and 77. Glandular cells from nonpathological Pap smears were used as controls. For each case, the means, medians, standard deviations, and the minimum and maximum values of both nuclear areas and perimeters of the cells of interest were calculated. The nuclear area analysis showed a 100% specificity in discriminating Normal from Altered cells (either Reactive or AGC), whereas the nuclear perimeter analysis showed a lower specificity (87.5%). Both nuclear area and perimeter variability analysis resulted in high specificity values in distinguishing Reactive cells from AGC. Therefore, a stepwise two-step algorithm using nuclear areas to discriminate Normal from Altered cells, and nuclear area variability to distinguish Reactive from AGC, allowed us to reliably classify the cells into these three categories. The morphometric analysis of nuclear area is a valuable and reliable aid in AGC diagnosis and standardization, easily integrable into common automatic algorithms.

Morphometric analysis of atypical glandular cells correctly classifies normal, reactive, and atypical cells in cervical smears

Perrella E;Bianchi A;Petitti T;Rabitti C;Crescenzi A.
2019-01-01

Abstract

The 2014 Bethesda System diagnostic criteria for atypical glandular cells (AGC) aid in the classification of atypical cells in cervical cytology. Anyway, AGC diagnosis remains challenging, due to low frequencies of this finding (approximately 0.5%-1% of Pap test results), abundance of AGC mimics, and significant interobserver variability. We developed an algorithm based on nuclear areas parameter that can help to differentiate AGC from Normal and Reactive glandular cells. Nuclear areas and perimeters were measured on 16 Pap smears with AGC and 18 with Reactive glandular cells of women aged between 30 and 77. Glandular cells from nonpathological Pap smears were used as controls. For each case, the means, medians, standard deviations, and the minimum and maximum values of both nuclear areas and perimeters of the cells of interest were calculated. The nuclear area analysis showed a 100% specificity in discriminating Normal from Altered cells (either Reactive or AGC), whereas the nuclear perimeter analysis showed a lower specificity (87.5%). Both nuclear area and perimeter variability analysis resulted in high specificity values in distinguishing Reactive cells from AGC. Therefore, a stepwise two-step algorithm using nuclear areas to discriminate Normal from Altered cells, and nuclear area variability to distinguish Reactive from AGC, allowed us to reliably classify the cells into these three categories. The morphometric analysis of nuclear area is a valuable and reliable aid in AGC diagnosis and standardization, easily integrable into common automatic algorithms.
2019
AGC; morphometry; pap test
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12610/4919
Citazioni
  • ???jsp.display-item.citation.pmc??? 0
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 1
social impact