DETECTION AND SUMMATION OF SQUAMOUS EPITHELIAL CELLS IN SPUTUM SLIDE IMAGES BY NUCLEUS DETECTION
Fatin Izzwani Azman, Kamarul Hawari Ghazali, Rosyati Hamid, Zeehaida Mohamed, Nur Shahida Nawi · 2013
Sputum with good quality is important to detect diseases. The presence of squamous epithelial cells (SEC) in sputum slide images is important to determine the quality of sputum. However the process is tedious especially in the existence of overlapping cells into single ones. Therefore, this paper proposes a system to detect overlapping cells by nucleus detection in order to count and sum the total of the SEC in the sputum slide images. Firstly, method has been proposed to segment cells by converting the RGB sputum image into HSV (Hue, Saturation, Value) color space and only consider the Value band from the HSV color space. Image multiplication is required in order to get the segmentation result. After the segmentation process, K-Means clustering technique is applied to determine each part of the cell; nucleus and the cytoplasm. After through some morphology process, the identified nucleus will be calculated. The approach is evaluated using 20 images containing hundreds of SEC and the proposed technique is capable to detect and count each SEC from overlapping SEC image. Therefore, the resultant image will be more useful for further analysis.