Method for segmenting cervix uteri liquid base cell image

Abstract

The invention discloses a method for segmenting a cervix uteri liquid base cell image. The method comprises the following steps of: 1, cell segmentation: performing illumination correction on an image, filtering noises and enhancing a cell boundary; and dividing the image into three types of cytoplasm, nucleus and background to eliminate a background class area; 2, nucleus segmentation: detecting centers of each of nucleuses in the cell area, estimating the approximate shapes of the nucleuses close to the centers of the nucleuses and correcting the shapes of the nucleuses to acquire an accurate boundary. The method is applied to cervix uteri cells and cell masses in a wide vision range, so that the phenomena of changed illumination, non-uniform dyeing and noise influence to a certain extent are avoided, and the cytoplasm, the single nucleus and adhesion nucleus can be segmented reliably and accurately.

Description

technical field [0001] The invention relates to cytopathology, computer vision and image processing methods, in particular to a method for automatic segmentation of cervical liquid-based cell images. Background technique [0002] Cervical cancer is one of the most threatening malignant tumors for women, and nearly 250,000 people die from cervical cancer every year in the world. In fact, cervical cancer is a curable cancer, as long as the cancer or HPV (human papillomavirus) infection can be detected early. At present, the most mature and effective method for clinical detection of early cervical cancer is liquid-based cytology combined with Pap staining. This preparation technique enables most cells to be spread evenly on the glass slide, thus facilitating the observation of cell nuclei and cytoplasm; this staining method enables the cells of different layers to display different colors, so that the cells of different layers can be differentiated to identify. It has been r...

AI Technical Summary

Problems solved by technology

However, the active contour model used by Harandi et al. to extract the cytoplasmic boundary is sensitive to noise. When there are impurities, air bubbles, inflammatory cells and other objects in the background, the active contour will stop converging and the cytoplasmic boundary cannot be obtained.

The traditional method of segmenting the nucleus is usually to binarize the image directly in the cytoplasmic region. This region-based segmentation method will cause many false positive targets (such as cytoplasmic overlapping parts), and it is difficult to extract when the nuclei are stained unevenly. Accurate nuclei borders, additionally no segmentation of cohesive nuclei

There are many technologies for the segmentation of cohesive nuclei, but these technologies have some shortcomings, such as: the curvature-based segmentation technology relies on the positioning of pit pairs, and when the pit detection is inaccurate, the segmentation performance is difficult to guarantee; the segmentation based on watershed Technology often leads to over-segmentation results; segmentation techniques based on distance maps cannot segment overlapping cells and cell groups; segmentation techniques based on mathematical morphology are very sensitive to the selection of structural elements and thresholds

Images