• Korean Journal of Radiology
• /
• 제24권4호
• /
• pp.294-304
• /
• 2023

Objective: We aimed to investigate whether image standardization using deep learning-based computed tomography (CT) image conversion would improve the performance of deep learning-based automated hepatic segmentation across various reconstruction methods. Materials and Methods: We collected contrast-enhanced dual-energy CT of the abdomen that was obtained using various reconstruction methods, including filtered back projection, iterative reconstruction, optimum contrast, and monoenergetic images with 40, 60, and 80 keV. A deep learning based image conversion algorithm was developed to standardize the CT images using 142 CT examinations (128 for training and 14 for tuning). A separate set of 43 CT examinations from 42 patients (mean age, 10.1 years) was used as the test data. A commercial software program (MEDIP PRO v2.0.0.0, MEDICALIP Co. Ltd.) based on 2D U-NET was used to create liver segmentation masks with liver volume. The original 80 keV images were used as the ground truth. We used the paired t-test to compare the segmentation performance in the Dice similarity coefficient (DSC) and difference ratio of the liver volume relative to the ground truth volume before and after image standardization. The concordance correlation coefficient (CCC) was used to assess the agreement between the segmented liver volume and ground-truth volume. Results: The original CT images showed variable and poor segmentation performances. The standardized images achieved significantly higher DSCs for liver segmentation than the original images (DSC [original, 5.40%-91.27%] vs. [standardized, 93.16%-96.74%], all P < 0.001). The difference ratio of liver volume also decreased significantly after image conversion (original, 9.84%-91.37% vs. standardized, 1.99%-4.41%). In all protocols, CCCs improved after image conversion (original, -0.006-0.964 vs. standardized, 0.990-0.998). Conclusion: Deep learning-based CT image standardization can improve the performance of automated hepatic segmentation using CT images reconstructed using various methods. Deep learning-based CT image conversion may have the potential to improve the generalizability of the segmentation network.

• 토지주택연구
• /
• 제2권1호
• /
• pp.69-78
• /
• 2011

본 연구는 디지털 이미지 처리 기법을 이용하여 골재입자의 형상을 분류한 것이다. 디지털 이미지 처리 기법은 골재의 표면거칠기 및 형상계수 등의 정량적인 분석에 적합한 것이다. 연구에 사용된 골재입자는 주문진 표준사, 해양골재 2종, 쇄석골재 2종 등이 이용되었다. 해변자갈I, II의 형상계수는 0.35~0.54, 쇄석I은 0.74의 범위로 박편 형태의 편평상으로 나타나는데 비해 주문진 표준사의 입자의 모양은 세장형으로 나타났다. 특히 해변자갈II는 해변자갈I과 달리 모든 시료보다 세장비가 작으므로 장축과 중축의 차이가 크고, 긴 입자의 형상특성을 보여주고 있다. 체가름 시험 및 실측을 통해 결정된 골재의 세장비 및 편평비와 디지털 이미지 처리 기법을 이용해 결정된 골재의 세장비 편평비는 거의 차이가 없는 것으로 나타났다. 또한, 디지털 이미지 처리 기법을 이용하여 2차원 이미지를 3차원으로 변환하기 위하여 형상변환계수 및 등가직경이론을 제안하였고, 이를 이용하여 수정입도분포곡선을 작성하였다. 4종의 골재에 대한 편평비는 각각 0.30, 0.36, 0.47 및 0.83으로 측정되었다. 형상변환계수는 각각 0.77, 0.78, 0.84 및 0.92로 결정되었다. 형상변환계수에 디지털이미지 처리기법에서 구한 입경을 곱하여 입자의 크기를 보정하였다. 보정된 입자크기를 이용하여 보정된 체적 및 중량을 구하여 수정입도곡선을 작성하였다.

• 한국산학기술학회논문지
• /
• 제22권4호
• /
• pp.183-194
• /
• 2021

본 연구에서는 SGR 모형과 AR 모형으로 부터 개발된 유형별 외래 및 입원 환산지수 분리 모형을 바탕으로 실제 건강보험 진료비 자료를 사용하여 2020년도 환산지수를 추정하였다. 또한 유형별 환산지수 계약 하에서 병원의 단일 환산지수 조정률과 외래 및 입원으로 분리된 복수의 환산지수 조정률을 산출하였다. 마지막으로 환산지수 분리 모형이 의료전달체계를 확립하기 위한 수단의 하나로 효과적으로 사용되기 위한 정책적 방안을 제안하였다. 본 연구의 주요 결론은 다음과 같다. 첫째, r=0.1에서 2020년도 병원의 단일 환산지수 조정률은 2.0%이며 병원의 환산지수를 분리하는 경우 외래 및 입원 환산지수 조정률은 각 각 2.2%, 2.3%로 나타났다. 그리고 조정계수를 활용하여 외래 및 입원 환산지수를 결합하는 방안을 제시하였다. 둘째, 의료전달체계 확립을 위한 조치의 하나로 제안된 가산율 조정 방식 대신에 가산율과 환산지수를 연동하여 운영하는 방안을 제안하였다. 셋째, 병원⦁의원 등 유형별로 외래환산지수, 입원환산지수, 조정계수를 도입하는 것에 추가해 진료량에 대한 목표관리를 가능하게 하는 모형 개발이 필요하다.

• 대한원격탐사학회지
• /
• 제27권6호
• /
• pp.729-741
• /
• 2011

Wind speeds have long been estimated from C-band VV-polarized SAR data by using the CMOD algorithms such as CMOD4, CMOD5, and CMOD_IFR2. Some SAR data with HH-polarization without any observations in VV-polarization mode should be converted to VV-polarized value in order to use the previous algorithms based on VV-polarized observation. To satisfy the necessity of polarization ratio (PR) for the conversion, we retrieved the conversion parameter from full-polarized SIR-C SAR image off the east coast of Korea. The polarization ratio for SIR-C SAR data was estimated to 0.47. To assess the accuracy of the polarization ratio coefficient, pseudo VV-polarized normalized radar cross section (NRCS) values were calculated and compared with the original VV-polarized ones. As a result, the estimated psudo values showed a good agreement with the original VV-polarized data with an root mean square error by 0.99 dB. We applied the psudo NRCS to the estimation of wind speeds based on the CMOD wind models. Comparison of the retrieved wind field with the ECMWF and NCEP/NCAR reanalysis wind data showed relatively small rms errors of 1.88 and 1.91 m/s, respectively. SIR-C HH-polarized SAR wind retrievals met the requirement of the scatterometer winds in overall. However, the polarization ratio coefficient revealed dependence on NRCS value, wind speed, and incident angle.

• Journal of Advanced Marine Engineering and Technology
• /
• 제39권2호
• /
• pp.143-151
• /
• 2015

Recently, tidal current energy conversion is a promising way to harness the power of tides in order to meet the growing demands of energy utilization. A new concept of tidal current energy conversion device, named Vane Tidal Turbine (VTT), is introduced in this study. VTT has several special features that are potentially more advantageous than the conventional tidal turbines, such as propeller type tidal turbines. The purpose of this study on VTT is to analyze the possibility of extracting the hydrokinetic energy of tidal current and converting it into electricity, and evaluate the performance of turbines for various numbers of blades (six, eight and twelve) using Computational Fluid Dynamics (CFD). At various tip-speed ratios (TSR), the six-bladed turbine obtains the highest power and torque coefficients, power efficiency is up to 28% at TSR = 1.89. Otherwise, the twelve blade design captures the smallest portion of available tidal current energy at all TSRs. However, by adding more blades, torque extracted from the rotor shaft of twelve-bladed turbine is more uniform due to the less interrupted generation of force for a period of time (one revolution).

• Nuclear Engineering and Technology
• /
• 제56권6호
• /
• pp.2190-2194
• /
• 2024

The study investigates the dose-dependent direct X-ray sensing characteristics of Brucinium benzilate (BB) and N-acetylglycine (NAG) organic crystals. BB and NAG were prepared as a slurry and deposited as a thick film on a patterned metal electrode. The X-ray induced photocurrent response was examined for various exposure doses using an intraoral pulsed 70 keV X-ray machine connected to a source meter. Subsequently, the morphological properties and thickness of the thick films were analyzed using scanning electron microscopy (SEM). At a photon energy of 70 keV, the attenuation coefficient values for NAG and BB crystals were determined to be approximately 0.181 and 0.178 cm²/g, respectively. The X-ray stopping power of the crystals was measured using a suniray-2 X-ray imaging system. To evaluate the responsiveness of the sensors, the photocurrent sensitivity and noise equivalent dose rate (NED) were calculated for both thick films. The findings demonstrated a noteworthy capability of sensing low doses (mGy), thereby suggesting the potential application of these organic materials in X-ray sensor development.

• Journal of Mechanical Science and Technology
• /
• 제19권9호
• /
• pp.1706-1710
• /
• 2005

In a case of computer simulation used for the verification of pneumatic system performance one of the main problems is that various parameters can be used to describe flow characteristics of the system components. The Standard ISO 6358 offers two parameters: the sonic conductance C and the critical static pressure ratio b, but the parameters can not be directly utilised in an analysis of a pneumatic system. In the standard analysis there is applied the airflow coefficient ${\mu}$, but it is not presented in the vendors' catalogues. In the paper the numerical algorithm for calculation of the airflow coefficient ${\mu}$. (which is required for computer simulation) as a function of sonic conductance C and a critical pressure ratio b (recommended by the standard) is presented. Additionally, because of the iterative character of the described algorithm, an artificial neural network approach to solve the problem is proposed.

• KIEE International Transaction on Electrical Machinery and Energy Conversion Systems
• /
• 제3B권3호
• /
• pp.115-121
• /
• 2003

This paper proposes a radiation model, which considers radiation transport as an important component in hot gas analysis. This radiation model is derived from combining the method of partial characteristics (MPC) with net emission coefficient (NEC), and it covers the drawbacks of existing models. Subsequently, using this proposed model, the arc-flow interaction in an arcing chamber can be efficiently computed. The arc is represented as an energy source term composed of ohmic heating and the radiation transport in the energy conservation equation. Ohmic heating term was computed by the electric field analysis within the conducting plasma region. Radiation transport was calculated by the proposed radiation model. Also, in this paper, radiation models were introduced and applied to the gas circuit breaker (GCB) model. Through simulation results, the efficiency of the proposed model was confirmed.

• 수산해양기술연구
• /
• 제55권4호
• /
• pp.411-418
• /
• 2019

In this study, to investigate the effect of physical and chemical properties of butanol on the engine performance and combustion characteristics, the coefficient of variations of IMEP (indicated mean effective pressure) and fuel conversion efficiency were obtained by measuring the combustion pressure and the fuel consumption quantity according to the engine load and the mixing ratio of diesel oil and butanol. In addition, the combustion pressure was analyzed to obtain the pressure increasing rate and heat release rate, and then the combustion temperature was calculated using a single zone combustion model. The experimental and analysis results of butanol blending oil were compared with the those of diesel oil under the similar operation conditions to determine the performance of the engine and combustion characteristics. As a result, the combustion stabilities of D.O. and butanol blending oil were good in this experimental range, and the indicated fuel conversion efficiency of butanol blending oil was slightly higher at low load but that of D.O. was higher above medium load. The premixed combustion period of D.O. was almost constant regardless of the load. As the load was lower and the butanol blending ratio was higher, the premixed combustion period of butanol blending oil was longer and the premixed combustion period was almost constant at high load regardless of butanol blending ratio. The average heat release rate was higher with increasing loads; especially as butanol blending ratio was increased at high load, the average heat release rate of butanol blending oil was higher than that of D.O. In addition, the calculated maximum. combustion temperature of butanol blending oil was higher than that of D.O. at all loads.

• 한국음향학회지
• /
• 제35권2호
• /
• pp.134-142
• /
• 2016

본 논문에서는 40 kHz 초음파 신호를 입 주변에 쏘고, 되돌아오는 초음파 신호를 이용해 음성신호를 합성하는 방법을 소개하고 성능을 평가하였다. 발성하고 있는 입주변에 초음파를 방사하게 되면, 입술, 턱, 뺨 등의 움직임으로 인한 변위로 도플러 현상이 발생하고, 이에 따라 반사 신호에는 본래의 주파수 성분과는 다른 도플러 주파수가 관찰되는데, 본 논문에서는 이러한 도플러 주파수를 이용하여 음성 파라메터를 추정하도록 하였다. 음성합성에 앞서서 초음파 도플러 신호와 음성 신호 간의 상관관계를 각 주파수 별로 분석하였으며, 이로부터 초음파 도플러 신호를 이용한 음성 신호의 합성 가능성을 살펴보았다. 변환에는 초음파 도플러의 정적, 동적 특성을 함께 반영한 특징 변수를 사용하였으며 결합-혼합 가우시안 기법을 이용하여 음성 파라메터로 변환하였다. 5명의 피 실험자를 이용한 음성 합성 실험에서 필터뱅크 에너지 값을 초음파신호의 특징변수로, LPC(Linear Predictive Coefficient) 켑스트럼 계수를 음성 변수로 사용하는 경우 가장 우수한 변환 성능을 나타내었다. 음성신호에서 추출한 여기신호를 이용하여 합성음을 생성하고, 이를 청취하였을 때 72.2 %의 평균 인식율이 얻어짐을 확인할 수 있었다.