Computer science is useful for detecting cervical cancer cells as early as possible.
REPUBLIKA.CO.ID, By Not Merlin
Computer science, as many people know, is a science that deals with programming and hardware. However, along with the development of computer science, now it has contributed a lot to almost all fields including the health sector.
Role computer science in the health sector, one of which is to be able to detect the presence of Pap Smear whose magnitude is between 5µm – 8µm, can also identify the part of the cytoplasm that is a component of Pap smear cells. With the development of computer science, it can assist pathologists in detecting cell abnormalities that may occur.
The term Pap Smear itself is a technical term for diagnosing diseases that threaten women so that it becomes a barometer of early detection of cervical cancer. Taking mucus with a spatula and placing it in a container called a preparation is what is done in the Pap Smear technique. Then the pathologist will examine it using a microscope.
Pap smear cells resemble cow’s eye eggs, where there are two important parts, namely the cell nucleus and cytoplasm. The nucleus of the cell is in the middle and the cytoplasm that surrounds the nucleus of the cell. The size and color of each of these cells affect the level of malignancy of cervical cancer. Therefore, the stage of finding the presence of cells will be the main step for the next process.
To obtain images of these cervical cells, a Logitech camera (Logitech HD C525 web cam) adapted to an optical microscope (Olympus CH20) was required. 40x magnification is used and the result is saved in JPEG format.
Learn from previous research, namely Pap Smear image research which is dominated by segmentation and classification topics, because these two topics are still a trend in the last 10 years.
The definition of an image (image) or what is called feature analysis in the cytoplasmic area in a Pap smear cell image becomes something interesting. This is due to image limitations and the complexity of morphological changes in the structural parts of cells. The analysis of features in the cytoplasmic area is important in the process of analyzing biomedical images. This is because the background has many obstacles (noise) and complex and poor cytoplasmic contrast.
There are many methods used to identify the size of the cytoplasm, segmentation methods for area on Pap Smear images that utilize the Gray Level Co-occurrence Matrix (GLCM) feature. The technique in GLCM is to perform texture analysis. The gray level of the Co-Occurrence Matrix has two important parameters, namely distance and direction. The GLCM feature extraction method is a matrix that describes the frequency of occurrence of two pixels.
This research was conducted to see how far the process of segmentation (grouping) of cytoplasmic color images using normal single cell images can produce features from texture and shape analysis. To analyze the shape of the cytoplasm, the RGB (Red, Green, Blue) color conversion method is used to convert HSV (Hue, Saturation, Value) color which produces a value metric and eccentricity. Then proceed with the process to determine the threshold image and calculate the area by changing the image threshold become an image binary.
As for texture analysis, analysis was carried out using the Gray Level Co-occurence Matrix (GLCM) method using the K-means method so that the parameters obtained contrast, correlation, energy, and homogenity. From this study, the results of segmentation (grouping) of normal Pap Smear single cell image samples were obtained so that it could be carried out properly and was able to obtain Metrix, Ecentricity, Contrast, Correlation and Energy features. Thus the area in the cytoplasm can be detected in order to detect the presence of cells cervical cancer as early as possible.
*) The author is the Vice Chancellor for Academic Affairs and a lecturer at Nusa Mandiri University, Information Systems Study Program.
– .
