• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.556-558
• /
• 2022

Diabetic retinopathy (DR) is one of the leading diseases causing vision loss. Early detection of this disease has a crucial role in protecting patients' eyes. Recent works have achieved impressive result when performing DR detection on fundus images using deep learning. In the deep learning-based approach, data augmentation has significant impact on the result. Recently, many data augmentation policies have been proposed and achieved state-of-the-art performance on different tasks. In this work, we compare effects of three data augmentation policies on DR grading in fundus images.

• Journal of Digital Convergence
• /
• v.12 no.10
• /
• pp.345-351
• /
• 2014

The purpose of this study is to offer more accurate information in whole spine examination of 3 spot DR through the comparative study about image distortion as making the flat phantom and measuring horizontal, vertical ratio and cobb angle of the virtual. We produced $H(40cm){\times}V(116cm){\times}D(2.3cm)$ flat acrylic phantom with lattice type of lead plate. We took projection respectively 3 times, total 9 times in each equipments using manufactured phantom as changing OFD to 6, 12, 18 cm. We measured a horizontal and vertical length of lead lattice and calculated the ratio. As appointing arbitrary points in the phantom and we measured cobb angle. The results of horizontal, vertical ratio measured CR type 0.98~1.01, scan DR type 0.96~0.97 and 3 spot DR 0.99~1.01. Cobb angle measured $52.5{\sim}53.3^{\circ}$, $52.1{\sim}54.3^{\circ}$ and $52.8{\sim}53.2^{\circ}$. Finally we can say that 3 spot DR method is an accurate method without any distortion in whole spine radiography.

• Journal of Biomedical Engineering Research
• /
• v.28 no.2
• /
• pp.199-204
• /
• 2007

In digital radiography, to improve the contrast of digital radiography image, the multi-scale nonlinear amplification algorithm based on unsharp masking is one of the major image enhancement algorithms. In this paper, we used the Laplacian pyramid to decompose a digital radiography(DR) image. In our simulation, the DR image was decomposed into seven layers and the coefficients of the each layer was amplified with nonlinear function. We also imported a noise containment algorithm to limit noise amplification. To enhance the contrast of image, we proposed a new adaptive non-linear gain amplification coefficients. As a result of having applied to some clinical data, a detail visibility was improved significantly without unacceptable noise boosting. Images that acquired with the proposed adaptive non-linear gain coefficients have shown superior quality to those that applied similar gain control method and expected to be accepted in the clinical applications.

• Journal of Korea Water Resources Association
• /
• v.47 no.12
• /
• pp.1227-1237
• /
• 2014

Dual-polarization can distinguish precipitation type and dual-polarization is provide not only meteorological phenomena in the atmosphere but also non-precipitation echoes. Therefore dual-polarization radar can improve radar estimates of rainfall. However polarimetric measurements by transmitting vertically vibration waves and horizontally vibrating waves simultaneously is contain systematic bias of the radar itself. Thus the calibration bias is necessary to improve quantitative precipitation estimation. In this study, the calibration bias of reflectivity (Z) and differential reflectivity ($Z_{DR}$) from the Bislsan dual-polarization radar is calculated using the 2-Dimensional Video Disdrometer (2DVD) data. And an improvement in rainfall estimation is investigated by applying derived calibration bias. A total of 33 rainfall cases occurring in Daegu from 2011 to 2012 were selected. As a results, the calibration bias of Z is about -0.3 to 5.5 dB, and $Z_{DR}$ is about -0.1 dB to 0.6 dB. In most cases, the Bislsan radar generally observes Z and $Z_{DR}$ variables lower than the simulated variables. Before and after calibration bias, compared estimated rainfall from the dual-polarization radar with AWS rain gauge in Daegu found that the mean bias has fallen by 1.69 to 1.54 mm/hr, and the RMSE has decreased by 2.54 to 1.73 mm/hr. And estimated rainfall comparing to the surface rain gauge as ground truth, rainfall estimation is improved about 7-61%.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.682-684
• /
• 2005

현재 설러 병원에서 PACS 시스템을 도입해 유용하게 사용하고 있다. 병원 진료에서 방사선 영상 정보는 중요한 위치를 차지한다. 방사선 영상은 관전압(KVP)과 관전류(mAs)로 방사선량을 조절한 후 인체에 조사하여 얻게 되는데, KVP와 mAs, 인체의 두께에 따라 영상의 질이 변하게 된다. 이와 같이 장비에서 촬영된 영상은 판독을 거처 진료에 이용되고 일정한 시간이 지나면 압축하여 보관하게 되는데, 압축율을 높게 적용할수록 저장장치에 대한 경제적인 효과는 크다. 이에 저자는 1) CR, DR 촬영 조건별 흉부 영상을 얻어 JPEG 2000 압축방식을 적용해 촬영조건이 압축영상에 미치는 영향을 평가하였고, 2) MOS(Mean opinion score) 평가를 통해 영상판독에 영향을 주지 않는 유효 압축율을 제시하였다.

• The Journal of the Korea Contents Association
• /
• v.19 no.12
• /
• pp.517-526
• /
• 2019

The purpose of this study is to find the optimal method for clinical application by analyzing image quality and radiation output according to parameter combination when using the Automatic Exposure Control (AEC). The experimental method combines 70, 81 kVp with sensitivity S200, S400, S800 and S1000 of the Automatic Exposure Control for Entrance Surface Dose (ESD), current volume, Signal to Noise Ratio (SNR), Contrast to Noise Ratio (CNR), Time-to-Radiation Dose Curve in abdomen and pelvis. And then, image quality and radiation output stability were evaluated. As a results, Entrance Surface Dose, current volume, Signal to Noise Ratio, Contrast to Noise Ratio decreased as the tube voltage and sensitivity were set higher. In addition, the higher tube voltage and sensitivity, the Time-to-Radiation Dose Curve showed a poor output stability. In conclusion, the higher the combination of tube voltage and sensitivity in the use of Automatic Exposure Control, the more problems can be seen in image quality and stability of the radiation output. Therefore, a relatively low combination of tube voltage and sensitivity showed that the image quality and radiation output stability could be optimized by minimizing the error range that would occur when the detector recognized a combination of parameters.

• The Journal of the Korea Contents Association
• /
• v.12 no.8
• /
• pp.233-239
• /
• 2012

This study was performed to find a method to decrease the radiation exposure of patients when obtaining anteroposterior pelvic images maintaining the quality of the image by using the copper filter built in the DR equipment now being used in our hospital. We measured the level of radiation by changing the filters from none filter to 0.1mmCu, 0.2mmCu, and 0.3mmCu and detected the organ level of radiation by PCXMC program. As a result, high levels of radiation were detected in the bladders and testicles and the levels were decreased as the thickness of the filter was increased and the amount of decrease was the highest in 0.1mmCu. And we asked for the expert opinions to 3 radiologists and as a result, only images obtained by 0.1mmCu filter out of all the images on which copper filters were used were accepted as the ones with diagnostic value same as none filter. At this time, the incident dose on the pelvic region was 0.895mGy which was smaller than the one in none filter by 47%. Therefore, using 0.1mmCu when obtaining anteroposterior pelvic images can effectively decrease the radiation exposure of patients.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.397-398
• /
• 2019

망막혈관 영상에서(retinal image) 혈관의 모양 또는 생성변화를 효과적으로 검진하기 위해서 망막혈관을 자동적으로 분리하는 영상분할 기법의 개발은 매우 중요한 사안이다. 이를 위해서 주로 망막혈관영상의 잡음을 억제하고 또한 혈관의 명암대비도(contrast)를 증가시키는 전처리 과정을 거쳐서 혈관의 국부적인 화소값의 변화, 방향성을 판별하여 혈관을 자동적으로 검출하는 방법들이 제시되어왔으며 최근에는 합성곱 신경망(CNN) 딥러닝 학습모델을 활용한 망막혈관 분리 알고리즘들이 제시되고 있다.

• Journal of the Korean Society of Radiology
• /
• v.2 no.4
• /
• pp.5-9
• /
• 2008

Antiscatter grids are widely used in radiography to remove scattered X-rays and thus improve the image contrast. However, the use of grids makes moir$\acute{e}$ artifact in the digital image, and this can be a critical reason for a mistaken diagnosis. In this paper, we examined that moire artifacts are how to relate with grid frequency, pixel pitch and grid rotation angle. To experiment we prepared 6 grids having different line frequencies (4.0 to 8.5lines/mm) and tested with a DR imager having a $139{\mu}m{\times}139{\mu}m$ pixel size. In the result of this experiment, we could get data about moir$\acute{e}$ artifact that could be make solution to remove the line artifact for the successful use of the grid in digital radiography. The acquired data and theory through this experiment, are expected to make contribution to the elimination of moir$\acute{e}$ artifact in the DR system.

• Journal of the Korean Society of Radiology
• /
• v.10 no.2
• /
• pp.117-124
• /
• 2016

The purpose of this study was tried to evaluate the filling sodium states inside the fuel rod of sodium-cooled fast reactor by digital medical X-ray. We used the diagnostic X-ray generators in digital radiography (DR). This study have found the optimal conditions by changing the effective focal spot size of X-ray tube and post-processing of the DR method with a tungsten edge plate in order to ensure excellent sharpness At this time, the sharpness and resolution were evaluated using the MTF (modulation transfer function). As a result, this study obtained a spatial resolution of 3.871 lp/mm (0.1 MTF), 3.290 lp/mm (0.5 MTF) when implemented the contrast strengthen post-processing in small focal spot. In this research, the result is able to evaluate the level of sodium inside the fuel rod by using the diagnostic X-ray generators in medical digital radiographic images.