Overlapping Cervical Cytology Image Segmentation Challenge

使用从子宫颈抹片获得的显微图像对重叠细胞进行自动检测和分割可以被认为是对宫颈细胞进行可靠的自动分析的主要障碍之一。子宫颈抹片是一种用于检测癌前和癌变过程的筛选测试，该测试由从子宫颈收集的细胞样本组成，这些细胞被涂抹在载玻片上，并在显微镜下进一步检查 (见左图)。影响巴氏涂片检查敏感性的主要因素是采样的细胞数量，这些细胞之间的重叠，细胞细胞质的对比度差以及粘液，血液和炎症细胞的存在。自动载玻片分析技术试图通过自动检测，分割和分类载玻片上存在的细胞来提高灵敏度和特异性。

在此挑战中，目标是从重叠的宫颈细胞学图像中提取单个细胞质和细胞核的边界。

Automatic detection and segmentation of overlapping cells from microscopic images acquired via Pap smears is considered one of the major barriers to reliable automated analysis of cervical cells. A Pap smear is a screening test used to detect precancerous and malignant processes, which consists of cell samples collected from the cervix, smeared onto glass slides, and further examined under a microscope (see the left image). The main factors affecting the sensitivity of Pap smear testing include the number of sampled cells, overlap between these cells, poor cytoplasmic contrast of cells, and the presence of mucus, blood, and inflammatory cells. Automated slide analysis technologies attempt to enhance sensitivity and specificity by automatically detecting, segmenting, and classifying the cells present on the slides.

In this challenge, the objective is to extract the boundaries of individual cytoplasm and cell nuclei from overlapping cervical cytological images.

搜集汇总

背景与挑战

背景概述

该数据集是一个重叠宫颈细胞学图像分割挑战，旨在从宫颈抹片显微图像中自动提取单个细胞的细胞质和细胞核边界，以提升宫颈癌筛查的自动分析可靠性。数据集由香港城市大学、昆士兰大学和阿德莱德大学于2013年12月16日发布。