• 한국의학물리학회지:의학물리
• /
• 제32권4호
• /
• pp.92-98
• /
• 2021

Purpose: This study automatically discriminates homogeneous and structure edge regions on computed tomography (CT) images, and it evaluates the noise level and edge preservation ratio (EPR) according to the different types of iterative reconstruction (IR). Methods: The dataset consisted of CT scans of 10 patients reconstructed with filtered back projection (FBP), statistical IR (iDose⁴), and iterative model-based reconstruction (IMR). Using the 10th and 85th percentiles of the structure coherence feature, homogeneous and structure edge regions were localized. The noise level was estimated using the averages of the standard deviations for five regions of interests (ROIs), and the EPR was calculated as the ratio of standard deviations between homogeneous and structural edge regions on subtraction CT between the FBP and IR. Results: The noise levels were 20.86±1.77 Hounsfield unit (HU), 13.50±1.14 HU, and 7.70±0.46 HU for FBP, iDose⁴, and IMR, respectively, which indicates that iDose⁴ and IMR could achieve noise reductions of approximately 35.17% and 62.97%, respectively. The EPR had values of 1.14±0.48 and 1.22±0.51 for iDose⁴ and IMR, respectively. Conclusions: The iDose⁴ and IMR algorithms can effectively reduce noise levels while maintaining the anatomical structure. This study suggested automated evaluation measurements of noise levels and EPRs, which are important aspects in CT image quality with patients' cases of FBP, iDose⁴, and IMR. We expect that the inclusion of other important image quality indices with a greater number of patients' cases will enable the establishment of integrated platforms for monitoring both CT image quality and radiation dose.

• 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.

• 대한디지털의료영상학회논문지
• /
• 제16권2호
• /
• pp.9-14
• /
• 2014

Purpose: The purpose of this study has attempted to evaluate and compare the image evaluation and exposure dose by respectively applying Filtered Back Projection(FBP), the existing test method, and Adaptive Statistical Iterative Reconstruction(ASIR) with different values of tube voltage during the Low Dose Computed Tomography(LDCT). Materials and Methods: With the image reconstruction method as basis, Chest Phantom was utilized with the FBP and ASIR set at 10%, 20% respectively, and the change of Tube Voltage (100kVp, 120kVp). For image evaluation, Back ground noise, Signal to Noise ratio(SNR) and Contrast to Noise ratio(CNR) were measured, and, for dose evaluation, CTDIvol and DLP were measured respectively. The statistical analysis was tested with SPSS(ver. 22.0), followed by ANOVA Test conducted after normality test and homogeneity test. (p<0.05). Results: In terms of image evaluation, there was no outstanding difference in Ascending Aorta(AA) SNR and Infraspinatus Muscle(IM) SNR with the different values of ASIR application(p<0.05), but a significant difference with the different amount of tube voltage(p>0.05). Also, there wasn't noticeable change in CNR with ASIR and different amount of Tube Voltage (p<0.05). However, in terms of dose evaluation, CTDIvol and DLP showed contrasting results(p<0.05). In terms of CTDIvol, the measured values with the same tube voltage of 120kVp were 2.6mGy with No-ASIR and 2.17mGy with 20%-ASIR respectively, decreased by 0.43mGy, and the values with 100kVp were 1.61mGy with No-ASIR and 1.34mGy with 20%-ASIR, decreased by 0.27mGy. In terms of DLP, the measured values with 120kVp were $103.21mGy{\cdot}cm$ with No-ASIR and $85.94mGy{\cdot}cm$ with 20%-ASIR, decreased by $17.27mGy{\cdot}cm$(about 16.7%), and the values with 100kVp were $63.84mGy{\cdot}cm$ with No-ASIR and $53.25mGy{\cdot}cm$ with 20%-ASIR, a decrease by $10.62mGy{\cdot}cm$(about 16.7%). Conclusion: At lower tube voltage, the rate of dose significantly decreased, but the negative effects on image evaluation was shown due to the increase of noise. For the future, through the result of the experiment, it is considered that the method above would be recommended for follow-up patients or those who get health checkup as long as there is no interference on the process of diagnosis due to the characteristics of Low Dose examination.

• 핵의학기술
• /
• 제21권1호
• /
• pp.15-22
• /
• 2017

본원에서 시행하는 Brain SPECT 검사는 $^{99m}Tc-ECD$ 또는 $^{99m}Tc-HMPAO$를 주사 한 후 뇌 영상을 얻어 뇌 관류상태를 평가하는 검사이다. 하지만 검사 중 일부 환자 상태가 불안정할 경우 움직임이 발생하여 재촬영이나 검사실패로 이어지는 경우가 발생 된다. 이에 현재의 Step and Shoot Mode(SSM)이 아닌 움직임이 발생되더라도 재구성을 통해 영상 구현이 가능한 Dynamic Continuous Mode(DCM)를 적용하여 환자의 재촬영과 피폭선량을 감소시키고 검사실에 운영 효율성을 높이고자함에 있다. Deluxe PET/SPECT Phantom과 Hoffman 3D Brain Phantom으로 Filtered Back Projection(FBJ)과 Iterated Reconstruction(IR)으로 재구성하여 영상을 구현하였다. 이미지를 가지고 핵의학과 5년이상의 임상경력이 있는 의사 5명과 방사선사 5명을 대상으로 리커트 5점 척도(Likert 5 Scale)와 블라인드 판독 테스트를 실시 하였다. 판독의 블라인드 테스트 결과 최소 DCM 3Repeat (30%)에서 7Repeat (50%)까지 판독에 영향을 주지 않는다고 답해 주었다. DCM으로 검사 시 환자 움직임이 발생되면 불필요한 부분을 제거하여 재촬영, 재주사의 감소를 가져올 수 있고, 장비 오류 시 영상을 재구성 후 구현 할 수 있어 검사실 운영 효율도 높을 수 있을 것으로 기대된다. 또한 SPECT검사뿐 만 아니라 SPECT/CT검사 에서도 활발한 연구가 적용 될 거라 기대 되며 마지막으로 실제 환자 적용은 환자 데이터의 충분한 수집 후 병원 판독 실정에 맞게 도입이 필요 하리라 사료된다.

• 한국정보디스플레이학회:학술대회논문집
• /
• 한국정보디스플레이학회 2006년도 6th International Meeting on Information Display
• /
• pp.421-421
• /
• 2006

The light and energy-efficiency of classical liquid crystal displays is notoriously poor due to the use of absorption-based linear polarisers and colour filters. For instance, the light efficiency of PVAL polarisers is typically between 40 and 45 % and the colour filters have a typical efficiency below 35 % which results in a total light and energy-efficiency of the display below 10 %. In the past, a variety of polarizers were developed with an enhanced efficiency in generating linearly polarized light. Typically, these polarizers are based on the polarisationselective reflection, scattering or refraction of light i.e. one polarisation direction of light is directly transmitted to the LCD/viewer and the other polarization direction of light is depolarised and recycled which results in a typical efficiency for generating linearly polarized light of 70-85 %. Also, special colour filters have been proposed based on chiral-nematic reactive mesogens which increase the efficiency of generating colour. Despite the enormous progress in this field, a need persists for improved methods for generating polarized light and colour based on low cost optical components with a high efficiency. Here, the use of holographic phase gratings is reported for the generation of polarized light and colour. The phase grating are recorded in a photopolymer which is coated onto a backor frontlight for LCDs. Typically the recording is performed in the transmisson mode or in the waveguiding mode and slanted phase gratings are generated with their refractive index modulation at an angle between 20o and 45o with the normal of the substrate. It is shown that phase gratings with a high refractive index modulation and a high efficiency can be generated by a proper selection of the photopolymer and illumination conditions. These phase gratings coupleout linearly polarized light with a high contrast (> 100) and the light is directed directly to the LCD/viewer without the need for redirection foils. Dependent on the type of phase grating, the different colours are coupled-out at a slightly different angle which potentially increases the efficiency of classical colour filters. Moreover, the phase gratings are completely transparent in direct view which opens the possibility to use them in frontlights for LCDs. Holographic polarization gratings posses a periodic pattern in the polarization state of light (and not in the intensity of light). A periodic pattern in the polarization direction of linearly polarized light is obtained upon interference of two circularly polarized laser beams. In the second part of the lecture, it is shown that these periodic polarization patterns can be recorded in a linear photo-polymerizable polymer (LPP) and that such an alignment layer induces a period rotation in the director of (reactive and non-reactive) liquid crystals. By a proper design, optical components can be produced with only first order diffraction and with a very high efficiency (>0.98). It is shown that these diffraction gratings are potentially useful in projection displays with a high brightness and energy efficiency

• Korean Chemical Engineering Research
• /
• 제53권2호
• /
• pp.236-242
• /
• 2015

고분자전해질 연료전지 스택의 성능 및 주요 운전 변수를 예측하기 위해 부분최소자승법과 인공신경망의 두 가지 데이터 기반 모델링 기법을 제시한다. 30 kW급 고분자전해질 연료전지 스택 실험으로부터 확보한 데이터를 사용하여 부분최소자승 및 인공신경망 모델들을 구성한 후 각 모델의 예측 성능 및 계산 시간을 비교하였다. 모델의 복잡성을 줄이기 위해 부분최소자승법에 기초한 VIP(Variable Importance on PLS Projections) 선정기준을 모델링 절차에 포함하여, 초기 입력변수의 집합으로부터 모델링에 필요한 입력변수들을 선정하였다. 모델링 결과, 인공신경망이 스택의 평균 셀전압과 캐소드(cathode) 출구 온도를 예측하는데 있어서, 부분최소자승법 보다 우수한 성능을 보였다. 그러나 부분최소자승법 또한 입력변수와 출력변수 간에 선형적 상관관계만을 모델링 할 수 있음에도 불구하고 비교적 만족할 만한 예측 성능을 나타냈다. 모델의 정확도와 계산속도의 요구조건에 따라 두 모델링 기법은 고분자전해질 연료전지의 설계 및 운전 분야의 성능 예측, 온라인 및 오프라인 최적화, 제어 및 이상 진단을 위해 적용될 수 있을 것으로 판단된다.

• Korean Journal of Radiology
• /
• 제19권6호
• /
• pp.1187-1195
• /
• 2018

Objective: To compare correlations between pulmonary function test (PFT) results and different reconstruction algorithms and to suggest the optimal reconstruction protocol for computed tomography (CT) quantification of low lung attenuation areas and airways in healthy individuals. Materials and Methods: A total of 259 subjects with normal PFT and chest CT results were included. CT scans were reconstructed using filtered back projection, hybrid-iterative reconstruction, and model-based IR (MIR). For quantitative analysis, the emphysema index (EI) and wall area percentage (WA%) were determined. Subgroup analysis according to smoking history was also performed. Results: The EIs of all the reconstruction algorithms correlated significantly with the forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) (all p < 0.001). The EI of MIR showed the strongest correlation with FEV1/FVC (r = -0.437). WA% showed a significant correlation with FEV1 in all the reconstruction algorithms (all p < 0.05) correlated significantly with FEV1/FVC for MIR only (p < 0.001). The WA% of MIR showed the strongest correlations with FEV1 (r = -0.205) and FEV1/FVC (r = -0.250). In subgroup analysis, the EI of MIR had the strongest correlation with PFT in both eversmoker and never-smoker subgroups, although there was no significant difference in the EI between the reconstruction algorithms. WA% of MIR showed a significantly thinner airway thickness than the other algorithms ($49.7{\pm}7.6$ in ever-smokers and $49.5{\pm}7.5$ in never-smokers, all p < 0.001), and also showed the strongest correlation with PFT in both ever-smoker and never-smoker subgroups. Conclusion: CT quantification of low lung attenuation areas and airways by means of MIR showed the strongest correlation with PFT results among the algorithms used, in normal subjects.

• Korean Journal of Radiology
• /
• 제19권6호
• /
• pp.1119-1129
• /
• 2018

Objective: To compare the image quality of low-tube-voltage and low-iodine-concentration-contrast-medium (LVLC) computed tomography urography (CTU) with iterative reconstruction (IR) with that of conventional CTU. Materials and Methods: This prospective, multi-institutional, randomized controlled trial was performed at 16 hospitals using CT scanners from various vendors. Patients were randomly assigned to the following groups: 1) the LVLC-CTU (80 kVp and 240 mgI/mL) with IR group and 2) the conventional CTU (120 kVp and 350 mgI/mL) with filtered-back projection group. The overall diagnostic acceptability, sharpness, and noise were assessed. Additionally, the mean attenuation, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and figure of merit (FOM) in the urinary tract were evaluated. Results: The study included 299 patients (LVLC-CTU group: 150 patients; conventional CTU group: 149 patients). The LVLC-CTU group had a significantly lower effective radiation dose ($5.73{\pm}4.04$ vs. $8.43{\pm}4.38mSv$) compared to the conventional CTU group. LVLC-CTU showed at least standard diagnostic acceptability (score ${\geq}3$), but it was non-inferior when compared to conventional CTU. The mean attenuation value, mean SNR, CNR, and FOM in all pre-defined segments of the urinary tract were significantly higher in the LVLC-CTU group than in the conventional CTU group. Conclusion: The diagnostic acceptability and quantitative image quality of LVLC-CTU with IR are not inferior to those of conventional CTU. Additionally, LVLC-CTU with IR is beneficial because both radiation exposure and total iodine load are reduced.

• Korean Journal of Radiology
• /
• 제22권11호
• /
• pp.1850-1857
• /
• 2021

Objective: The purpose of this study was to assess whether a deep learning (DL) algorithm could enable simultaneous noise reduction and edge sharpening in low-dose lumbar spine CT. Materials and Methods: This retrospective study included 52 patients (26 male and 26 female; median age, 60.5 years) who had undergone CT-guided lumbar bone biopsy between October 2015 and April 2020. Initial 100-mAs survey images and 50-mAs intraprocedural images were reconstructed by filtered back projection. Denoising was performed using a vendor-agnostic DL model (ClariCT.AI^TM, ClariPI) for the 50-mAS images, and the 50-mAs, denoised 50-mAs, and 100-mAs CT images were compared. Noise, signal-to-noise ratio (SNR), and edge rise distance (ERD) for image sharpness were measured. The data were summarized as the mean ± standard deviation for these parameters. Two musculoskeletal radiologists assessed the visibility of the normal anatomical structures. Results: Noise was lower in the denoised 50-mAs images (36.38 ± 7.03 Hounsfield unit [HU]) than the 50-mAs (93.33 ± 25.36 HU) and 100-mAs (63.33 ± 16.09 HU) images (p < 0.001). The SNRs for the images in descending order were as follows: denoised 50-mAs (1.46 ± 0.54), 100-mAs (0.99 ± 0.34), and 50-mAs (0.58 ± 0.18) images (p < 0.001). The denoised 50-mAs images had better edge sharpness than the 100-mAs images at the vertebral body (ERD; 0.94 ± 0.2 mm vs. 1.05 ± 0.24 mm, p = 0.036) and the psoas (ERD; 0.42 ± 0.09 mm vs. 0.50 ± 0.12 mm, p = 0.002). The denoised 50-mAs images significantly improved the visualization of the normal anatomical structures (p < 0.001). Conclusion: DL-based reconstruction may enable simultaneous noise reduction and improvement in image quality with the preservation of edge sharpness on low-dose lumbar spine CT. Investigations on further radiation dose reduction and the clinical applicability of this technique are warranted.

• 대한영상의학회지
• /
• 제84권1호
• /
• pp.240-252
• /
• 2023

목적 소아 환자에서 두부 컴퓨터단층촬영(이하 CT)에 대한 딥러닝 이미지 재구성(deep learning image reconstruction; 이하 DLIR; TrueFidelity; GE Healthcare, Milwaukee, WI, USA)의 효과를 평가하고자 한다. 대상과 방법 총 126개의 소아 두부 CT 이미지를 수집했으며, adaptive statistical iterative reconstruction (이하 ASiR)-V를 사용한 반복적 재구성 및 세 가지 수준의 DLIR을 사용한 재구성을 시행하였다. 각 이미지 세트 그룹은 환자의 연령에 따라 4개의 그룹으로 구분하였으며 각 연령군의 임상 및 방사선량 관련 데이터를 검토하였다. 양적 매개 변수에는 signal to noise ratio (이하 SNR) 및 contrast to noise ratio (이하 CNR)가 포함되었으며 질적 매개 변수로 영상의 잡음(noise), 회백질의 구분 정도, 선명도, 인공물 및 수용 가능성(acceptability), 영상의 질감이 포함되었고 이에 대한 평가와 비교를 시행하였다. 결과 모든 연령 그룹의 모든 수준의 SNR 및 CNR은 높은 수준의 DLIR 사용 시 증가하였다. ASiR-V와 비교했을 때 높은 수준의 DLIR은 SNR 및 CNR이 개선되었다(p < 0.05). 그리고 DLIR의 수준이 증가될수록 순차적인 잡음 감소, 회백질 구분 개선, 선명도 개선이 나타났다. 이러한 변수들에서 높은 수준의 DLIR 사용 시 ASiR-V와 유사한 정도의 수치가 측정되었다. 인공물과 수용 가능성의 경우에 적용된 DLIR 수준 간에 큰 차이를 보이지 않았다. 결론 소아 두부 CT에 고수준 DLIR을 적용하면 영상의 노이즈를 줄일 수 있으나 인공물 처리에 대한 개선이 필요하다.