A contour property based approach to segment nuclei in cervical cytology images

Abstract

Background Segmentation of nuclei in cervical cytology pap smear images is a crucial stage in automated cervical cancer screening. The task itself is challenging due to the presence of cervical cells with spurious edges, overlapping cells, neutrophils, and artifacts. Methods After the initial preprocessing steps of adaptive thresholding, in our approach, the image passes through a convolution filter to filter out some noise. Then, contours from the resultant image are filtered by their distinctive contour properties followed by a nucleus size recovery procedure based on contour average intensity value. Results We evaluate our method on a public (benchmark) dataset collected from ISBI and also a private real dataset. The results show that our algorithm outperforms other state-of-the-art methods in nucleus segmentation on the ISBI dataset with a precision of 0.978 and recall of 0.933. A promising precision of 0.770 and a formidable recall of 0.886 on the private real dataset indicate that our algorithm can effectively detect and segment nuclei on real cervical cytology images. Tuning various parameters, the precision could be increased to as high as 0.949 with an acceptable decrease of recall to 0.759. Our method also managed an Aggregated Jaccard Index of 0.681 outperforming other state-of-the-art methods on the real dataset. Conclusion We have proposed a contour property-based approach for segmentation of nuclei. Our algorithm has several tunable parameters and is flexible enough to adapt to real practical scenarios and requirements.

Edge Enhancement and Detection Approach on Cervical Cytology Images

“…The findings demonstrate that the method has a higher detection accuracy than the methods that were compared. The finding is consistent with the study by Hoque et al, [11] which in the method proposed, the image goes through a convolution filter after the initial adaptive thresholding pre-processing steps to remove some noise. Then, a nucleus size recovery process based on the average intensity value of the contours, the contours from the resulting image are filtered according to their individual contour properties.…”

“…Segmentation of nuclei in cervical cytology pap smear images is a crucial stage in automated cervical cancer screening. Segmentation also become difficult due to the presence of cervical cells with spurious edges, overlapping cells, neutrophils, and artifacts [11].…”

Pap Smear Image Analysis Based on Nucleus Segmentation and Deep Learning – A Recent Review

“…Consequently, Hoque et al, [20] presented a strategy in which the image runs via a convolution filter to filter out certain noise after the initial pre-processing stages of adaptive thresholding. The contours in the resulting image are subsequently filtered, relying on their different contour qualities, and a nucleus size recovery technique relying on the contour average intensity value is performed.…”

“…The accuracy of the proposed techniques of nuclei segmentation by Devi and Thirumurugan [4], Hoque et al, [20], and Jia et al, [22] are 97.8%, 91.3%, and 96.31%, respectively. Together, these results provide important insights into the nuclei segmentation technique as one of the developing techniques in Pap smear image analysis.…”

“…An important element in automated cervical cancer screening is nuclei segmentation in cytology Pap smear images. The work is arduous because cervical cells have artifacts, neutrophils, overlapping cells, and spurious edges [20,21]. Precise cell segmentation, especially nuclei, is critical to the diagnostic and analysis procedure.…”