UNAIR NEWS – Digital imaging technology has been widely used in the medical field and diagnoses biological image data to guide doctors and find out the patient’s condition. One medical imaging technique that can describe conditions in the human body, such as computed tomography (CT).
Regarding to this, Faculty of Vocational Studies Lecturer, Riky Tri Yunardi, S.T., M.T., conducted a study to measure the quality of abdominal CT scan images. According to him, to optimize the quality of CT images, especially tomography of the abdominal kernel (abdomen) can be applied through contrast-enhanced value analysis.
“In this study, the measurement of image quality from the abdominal CT scan using contrast values aims to evaluate the resulting image,” he explained.
He explained that the method used for quantitative measurements through image quality parameters could be evaluated in three aspects namely the peak signal to noise ratio (PSNR), root means square error (RMSE), and maximum absolute error (MAE).
“As an addition to the analysis data, a subjective evaluation is also carried out by one radiologist to determine the selection of the appropriate kernel,” he said.
Furthermore, for the process of measuring image quality, the steps taken include analyzing and comparing the value of improving image quality in digital images using quantitative measurement methods. As input, he explained, CT scan images that have been filtered using contrast-enhanced so that the results will be smoother (final image) compared to the initial image.
“This image is measured by the density parameter of noise using the mean square error equation and to provide the value of the average relative weight of the error using the root mean square error (RMSE). This measurement is done with a grayscale image value that has the same size in the area of pixels, “he said.
In the end, the analysis relating to the strength of the ratio of noise use the peak signal to noise ratio (PSNR). The results of calculations are expressed on a logarithmic decibel scale. It can be interpreted as a comparison between the maximum value of the signal measured by the amount of noise that affects the image.
Author: Nuri Hermawan
Editor: Khefti Al Mawalia
Detail of this research:
https://ieeexplore.ieee.org/document/8834645
Yunardi, Riky Tri, Qurratul Istiqomah, and Risalatul Latifah. “Contrast-enhanced Based on Abdominal Kernels for CT Image Noise Reduction.” In 2019 International Conference of Artificial Intelligence and Information Technology (ICAIIT), pp. 293-296. IEEE, 2019.
