Morphometric analysis of endometrial cells in liquid‐based cervical cytology samples

Abstract

We conclude that the morphometric parameters of the nuclear area and nuclear perimeter are helpful in differentiating between benign, atypical and malignant endometrial nuclei even in liquid-based cervical cytology samples.

Role of fractal dimension in distinguishing benign from malignant endometrial clusters in liquid‐based cervical samples

“…Although the distinction between benign and malignant endometrial cell cluster can be made in a considerable number of cases on the basis of cytomorphology alone, there are cases where this distinction is challenging. Recently, some authors have studied the diagnostic utility of certain newer advanced methods for this purpose; so far most of these have been used as an adjunct to routine cytomorphology, on either cervical samples or direct endometrial samples, for differentiating malignant from benign endometrial cells with reactive atypia, especially in difficult cases …”

“…As a consequence of this, it often becomes challenging for the cytopathologists to differentiate between benign and malignant endometrial cells . Some studies have been conducted to evaluate the potential role of additional parameters such as markers of chromosomal instability and nuclear morphometric measurements for accurate distinction of benign from malignant endometrial cells on cervical samples . Kiraz et al, in their study on women with endometrial hyperplasia and carcinoma found that micronucleus ratio was significantly higher in carcinoma group compared to the hyperplasia group, thereby concluding that micronucleus assay can be used as an adjunct in endometrial cancer screening.…”

“…Gupta et al, also concluded that micronucleus scoring on cervical smears can be used as a potential adjunct to routine cytomorphology for diagnosis of endometrial adenocarcinoma. In another study involving multiple morphometric parameters, authors concluded that nuclear area and nuclear perimeter can be used to differentiate benign from the malignant endometrial cells in liquid‐based cervical cytology samples …”

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

“…We then validated this cut-off on our data using the following rule: if the CV of the perimeters was higher or equal to 0.15, the sample would have been classified as AGC, otherwise, as Reactive ( [29][30][31] Due to the rarity of this condition, we could not validate the performance of this algorithm on an independent dataset. The value of 0.15 was considered as a cut-off score for the distinction between Reactive cells and AGC.…”

“…Of note, this value is significantly higher than the 95% confidence interval of Reactive cells, which is equal to 0.12. We then validated this cut-off on our data using the following rule: if the CV of the perimeters was higher or equal to 0.15, the sample would have been classified as AGC, otherwise, as Reactive ( [29][30][31] Due to the rarity of this condition, we could not validate the performance of this algorithm on an independent dataset. Another limitation of this algorithm is that it is not fully automated, because it relies on the direct measurement of the morphometric parameters by an operator.…”