• Journal of the Korean Society of Radiology
• /
• v.17 no.5
• /
• pp.709-715
• /
• 2023

In this study, the performance evaluation of image quality for noise reduction was implemented using the U-net deep learning architecture in computed tomography (CT) images. In order to generate input data, the Gaussian noise was applied to ground truth (GT) data, and datasets were consisted of 8:1:1 ratio of train, validation, and test sets among 1300 CT images. The Adagrad, Adam, and AdamW were used as optimizer function, and 10, 50 and 100 times for number of epochs were applied. In addition, learning rates of 0.01, 0.001, and 0.0001 were applied using the U-net deep learning model to compare the output image quality. To analyze the quantitative values, the peak signal to noise ratio (PSNR) and coefficient of variation (COV) were calculated. Based on the results, deep learning model was useful for noise reduction. We suggested that optimized hyper parameters for noise reduction in CT images were AdamW optimizer function, 100 times number of epochs and 0.0001 learning rates.

• Journal of the Korean Society of Radiology
• /
• v.18 no.1
• /
• pp.21-26
• /
• 2024

The one of the famous deep learning models for object detection task is you only look once version 5 (YOLOv5) framework based on the one stage architecture. In addition, YOLOv5 model indicated high performance for accurate lesion detection using the bottleneck CSP layer and skip connection function. The purpose of this study was to evaluate the performance of YOLOv5 framework according to various hyperparameters in position emission tomogrpahy (PET) phantom images. The dataset was obtained from QIN PET segmentation challenge in 500 slices. We set the bounding box to generate ground truth dataset using labelImg software. The hyperparameters for network train were applied by changing optimization function (SDG, Adam, and AdamW), activation function (SiLU, LeakyRelu, Mish, and Hardwish), and YOLOv5 model size (nano, small, large, and xlarge). The intersection over union (IOU) method was used for performance evaluation. As a results, the condition of outstanding performance is to apply AdamW, Hardwish, and nano size for optimization function, activation function and model version, respectively. In conclusion, we confirmed the usefulness of YOLOv5 network for object detection performance in nuclear medicine images.

• Journal of the Korean Society of Radiology
• /
• v.18 no.2
• /
• pp.93-99
• /
• 2024

The digital medical imaging, especially, computed tomography (CT), should necessarily be considered in terms of noise distribution caused by converting to X-ray photon to digital imaging signal. Recently, the denoising technique based on deep learning architecture is increasingly used in the medical imaging field. Here, we evaluated noise reduction effect according to various noise types based on the U-net deep learning model in the lung CT images. The input data for deep learning was generated by applying Gaussian noise, Poisson noise, salt and pepper noise and speckle noise from the ground truth (GT) image. In particular, two types of Gaussian noise input data were applied with standard deviation values of 30 and 50. There are applied hyper-parameters, which were Adam as optimizer function, 100 as epochs, and 0.0001 as learning rate, respectively. To analyze the quantitative values, the mean square error (MSE), the peak signal to noise ratio (PSNR) and coefficient of variation (COV) were calculated. According to the results, it was confirmed that the U-net model was effective for noise reduction all of the set conditions in this study. Especially, it showed the best performance in Gaussian noise.

• Journal of the Korean Society of Radiology
• /
• v.18 no.2
• /
• pp.83-91
• /
• 2024

The purpose of this study was to determine the doppler fluid effects of carbonated water (CBW) in a bi-directional flow phantom model. A bi-directional flow phantom model was chosen to realize arterial and venous flow, and the structure of the inner and outer tanks allowed for fluid circulation and also made the size of the phantom small. Carbonated water (CBW), salt fluid (SAF), sugar fluid (SUF), and distilled water (DW) were used as fluids, and ultrasound scans were performed at depths of 1.5 cm and 3.0 cm from the surface of the tank, using B-mode and color Doppler effects. All fluids tested showed color Doppler effects, but CBW had the highest doppler shift and the least variation with depth. In conclusion, we determined that CBW was the most suitable fluid to be used as a doppler fluid and confirmed that the bubbles dissolved in CBW act as doppler scatterers, just like red blood cells inside human blood. Therefore, it is possible that CBW can be used as a blood-mimicking fluid in doppler ultrasound phantoms through further research, and this study will provide basic data.

• Journal of the Korean Society of Radiology
• /
• v.18 no.3
• /
• pp.275-282
• /
• 2024

This study analyzed the factors that influence the development of gynecomastia by comparing blood test results and lifestyle between patients diagnosed with gynecomastia and normal control group. Independent sample t-test and chi-square tests were conducted to examine the mean differences among risk factors, and logistic regression analysis was used to calculate the relative risk ratio of various predictors influencing the development of gynecomastia. Age, HGB, HCT, TC, glucose, ALT, alcohol consumption, hair loss medication and medication for benign prostatic hyperplasia were identified as significant variables. Notably, the use of hair loss medication showed the highest univariate odds ratio of 15.79, while HGB decreased by a factor of 0.37. These results suggest that these could play a significant role in the onset of gynecomastia and could serve as foundational data for future diagnosis and management of the condition.

• Journal of the Korean Society of Radiology
• /
• v.18 no.5
• /
• pp.419-427
• /
• 2024

The objective of this study is to analyze and evaluate the diagnostic performance and utility of elastography and artificial intelligence program in distinguishing between benign and malignant thyroid nodule. In a hospital outpatient clinic, we performed thyroid ultrasound from January 2023 to June 2024, and retrospectively analyzed 126 patients who performed elastography, S-detect, and fine needle aspiration cytology(FNAC) because nodules were found. The analysis of differences based on cytology results showed statistically significant differences in age, nodule size, echogenicity, nodule orientation, margins, shape, presence of calcification, posterior shadowing, K-TIRADS ultrasound interpretation, S-detect results and elasticity contrast index. The ROC curve analysis determined a cut off value for the elasticity contrast index at 2.32, with diagnostic concordance rates of 0.66 for expert interpretation, 0.49 for S-detect, and 0.67 for the elasticity contrast index, indicating superior diagnostic performance with elastography. Thus, elastography may ge used as an adjunct tool to minimize unnecessary repeat examinations and the frequency of tissue biopsies in the diagnosis of thyroid nodules.

• Biomedical Science Letters
• /
• v.13 no.3
• /
• pp.207-212
• /
• 2007

The purpose of this preliminary study is to improve the efficiency of gene transfer of nonviral plasmid DNA by in vivo electroporation in ischemic hindlimb muscle, tibialis anterior. Hindlimb ischemic model was aseptically made by excision of left femoral artery. Each $50\;{\mu}g$ of pEGFP-C1 and pGL3-control in $100\;{\mu}l$ 0.9% NaCl was injected in tibialis anterior muscle. In vivo electroporation was applied on the same site with 10 mm-distance 2 needle array electrodes and ECM830. In 3 groups of normal rat with different electric field strength 0, 200 and 800 V/cm, the expression of pEGFP-C1 was comparatively evaluated. In 8 groups of normal rats, the expression of pGL3-control was evaluated in 0, 40, 50, 80, 100, 150, 200 and 300 V/cm of electric field strength. In 5 groups of ischemic models, the expression of pGL3-control was analyzed on 0, 4, 7, 10 and 14 days elapsed after making ischemic models. In 9 groups of ischemic rats, the expression of pGL3-control was analyzed in the electric field strength 0, 60, 70, 80, 100, 150, 200, 250 and 300 V/cm. GFP expressions in normal tibialis anterior were high in the extent and degree in order of electric field strength of 200, 800 and 0 V/cm. Luciferase value was highest in $50{\sim}100\;V/cm$ electric field strength. In the case of ischemic models, luciferase expression was significantly increasing in the order elapsed time after making the model. The degree of luciferase expression was higher in cases of application of in vivo electroporation than in that of non-application and was highest in $100{\sim}150\;V/cm$. In conclusion, in vivo electroporation is effective in transfer and expression of plasmid DNA in normal and ischemic tibialis anterior of rat.

• Imaging Science in Dentistry
• /
• v.43 no.3
• /
• pp.179-184
• /
• 2013

Purpose: X-ray exposure should be clinically justified and each exposure should be expected to give patients benefits. Since dental radiographic examination is one of the most frequent radiological procedures, radiation hazard becomes an important public health concern. The purpose of this study was to investigate the attitude of Korean dentists about radiation safety and use of criteria for selecting the frequency and type of radiographic examinations. Materials and Methods: The study included 267 Korean dentists. Five questions related to radiation safety were asked of each of them. These questions were about factors associated with radiation protection of patients and operators including the use of radiographic selection criteria for intraoral radiographic procedures. Results: The frequency of prescription of routine radiographic examination (an example is a panoramic radiograph for screening process for occult disease) was 34.1%, while that of selective radiography was 64.0%. Dentists' discussion of radiation risk and benefit with patients was infrequent. More than half of the operators held the image receptor by themselves during intraoral radiographic examinations. Lead apron/thyroid collars for patient protection were used by fewer than 22% of dental offices. Rectangular collimation was utilized by fewer than 15% of dental offices. Conclusion: The majority of Korean dentists in the study did not practice radiation protection procedures which would be required to minimize exposure to unnecessary radiation for patients and dental professionals. Mandatory continuing professional education in radiation safety and development of Korean radiographic selection criteria is recommended.

• Journal of the Korean Society of Radiology
• /
• v.14 no.6
• /
• pp.827-832
• /
• 2020

X-ray equipment, which is frequently used in radiology and treatment, is the most common and most used equipment in clinical practice. Equipment must provide accurate information to patients through continuous quality control. In case of manual quality control measurement, reproducibility may be poor and there may be a problem with reliability of evaluation results. In this study, an automated program was developed and attempted to measure how much the central ray between the focus of the X-ray tube and the variable aperture of the diagnostic X-ray generator used in clinical practice coincides. As a result of the experiment, it succeeded in calculating the coordinates of the two center points, and the distance between the two points was calculated in pixels and applied to the judgment and the automatic judgment value for whether the center line coincidence is within the normal angle or the abnormal angle is presented. The results of this study are considered to be very helpful in the quality control of the X-ray apparatus.

• Journal of the Korean Society of Radiology
• /
• v.16 no.3
• /
• pp.343-349
• /
• 2022

We intend to use topic modeling to identify radiation-themed papers published from 1989 to 2022 and analyze the relevance and weight between topics. This study analyzed topics derived from national subjects for 717 papers published until recently in 2022 to contribute to the revitalization of research in the field of radiation. Through text mining, overall research trends on the subject distribution of the study were analyzed, and five topics were derived through topic modeling. First, among the papers to be analyzed, a total of 1,675 words were frequency-analyzed through the preprocessing process of key words in a total of 717 papers centered on keywords. Second, as a result of analyzing topics based on the association of constituent words for five topics, it was found that studies focused on minimizing dose in the range that does not degrade image quality in the fields of radiation, image, CT clinical. In addition, it was found that various studies were mainly conducted in the MRI, and the study of ultrasound in various areas of disease analysis was actively attempted.