• Korean Journal of Radiology
• /
• 제22권3호
• /
• pp.476-488
• /
• 2021

Objective: We aimed to develop a deep neural network for segmenting lung parenchyma with extensive pathological conditions on non-contrast chest computed tomography (CT) images. Materials and Methods: Thin-section non-contrast chest CT images from 203 patients (115 males, 88 females; age range, 31-89 years) between January 2017 and May 2017 were included in the study, of which 150 cases had extensive lung parenchymal disease involving more than 40% of the parenchymal area. Parenchymal diseases included interstitial lung disease (ILD), emphysema, nontuberculous mycobacterial lung disease, tuberculous destroyed lung, pneumonia, lung cancer, and other diseases. Five experienced radiologists manually drew the margin of the lungs, slice by slice, on CT images. The dataset used to develop the network consisted of 157 cases for training, 20 cases for development, and 26 cases for internal validation. Two-dimensional (2D) U-Net and three-dimensional (3D) U-Net models were used for the task. The network was trained to segment the lung parenchyma as a whole and segment the right and left lung separately. The University Hospitals of Geneva ILD dataset, which contained high-resolution CT images of ILD, was used for external validation. Results: The Dice similarity coefficients for internal validation were 99.6 ± 0.3% (2D U-Net whole lung model), 99.5 ± 0.3% (2D U-Net separate lung model), 99.4 ± 0.5% (3D U-Net whole lung model), and 99.4 ± 0.5% (3D U-Net separate lung model). The Dice similarity coefficients for the external validation dataset were 98.4 ± 1.0% (2D U-Net whole lung model) and 98.4 ± 1.0% (2D U-Net separate lung model). In 31 cases, where the extent of ILD was larger than 75% of the lung parenchymal area, the Dice similarity coefficients were 97.9 ± 1.3% (2D U-Net whole lung model) and 98.0 ± 1.2% (2D U-Net separate lung model). Conclusion: The deep neural network achieved excellent performance in automatically delineating the boundaries of lung parenchyma with extensive pathological conditions on non-contrast chest CT images.

• ETRI Journal
• /
• 제33권3호
• /
• pp.374-381
• /
• 2011

A simple and effective compression method is proposed for multiple-scan testing. For a given test set, each test pattern is compressed from the view of slices. An encoding table exploiting seven types of frequently-occurring pattern is used. Compression is then achieved by mapping slice data into codewords. The decompression logic is small and easy to implement. It is also applicable to schemes adopting a single-scan chain. Experimental results show this method can achieve good compression effect.

• 한국멀티미디어학회논문지
• /
• 제23권5호
• /
• pp.625-632
• /
• 2020

In the diagnosis of lung cancer, the tumor size is measured by the longest diameter of the tumor in the entire slice of the CT. In order to accurately estimate the size of the tumor, it is better to measure the volume, but there are some limitations in calculating the volume in the clinic. In this study, we propose an algorithm to segment lung cancer by applying a custom loss function that combines focal loss and dice loss to a U-Net model that shows high performance in segmentation problems in chest CT images. The combination of values of the various parameters in custom loss function was compared to the results of the model learned. The purposed loss function showed F1 score of 88.77%, precision of 87.31%, recall of 90.30% and average precision of 0.827 at α=0.25, γ=4, β=0.7. The performance of the proposed custom loss function showed good performance in lung cancer segmentation.

• 대한디지털의료영상학회논문지
• /
• 제17권1호
• /
• pp.13-18
• /
• 2015

Purpose : Skip the repetitive HRCT axial scan in order to reduce the exposure of patients during chest HRCT scan, Helical Scan Data into a reconstructed image, and exposure of the patient change and visually evaluate the usefulness of the HRCT images. Materials and method : Patients were enrolled in the survey are 50 people who underwent chest CT scans of patients who presented to the hospital from January 2015 to March 2015. 50 people surveyed 22 people men and 28 people women people showed an average distribution of 30 to 80 years age was 48 years. 50 patients to Somatom Sensation 64 ch (Siemens) model with 120 kVp tube voltage to a reference mAs tube current to mAs (Care dose, Siemens) as a whole, including the lungs and the chest CT scan was performed. Scan upon each patient CARE dose 4D (Automatic exposure control, Siemens Medical Solution Erlangen, Germany) was to maintain the proper radiation dose scan every cross-section through a device that automatically adjusts the tube current of. CT scan is the rotation time of the Tube slice collimation, slice width 0.6 mm, pitch factor was made under the terms of 1.4. CT scan obtained after the raw data (raw data) to the upper surface of the axial images and coronal images for each slice thickness 1 mm, 5 mm intervals in the high spatial frequency calculation method (hight spatial resolution algorithm, B60 sharp) was the use of the lung window center -500 HU, windows were reconstructed into images in the interval -1000 HU to see. Result : 1. Measure the total value of DLP 50 patients who proceed to chest CT group A (Helical Scan after scan performed with HRCT) and group B (Helical Scan after the HR image reconstruction to the original data) compared with the group divided, analysis As a result of the age, but show little difference for each age group it had a decreased average dose of about 9%. 2. A Radiation read the results of the two Radiologist and a doctor upper lobe and middle lobe of the lung takes effect the visual evaluation is not a big difference between the two images both, depending on the age of the patient, especially if the blood vessels of the lower lobe (A: 3.4, B: 4.6) and bronchi(A: 3.8, B4.7) image shake caused by breathing in anxiety (blurring lead) to the original data (raw data) showed that the reconstructed image is been more useful in diagnostic terms. Conclusion : Scan was confirmed a continuous, rapid motion video to get Helical scan is much lower lobe lung reduction in visual blurring, Helical scan data to not repeat the examination by obtaining HRCT images reorganization reduced the exposure of the patient.

• 대한의용생체공학회:의공학회지
• /
• 제10권1호
• /
• pp.1-10
• /
• 1989

An automatic lung nodule detection algorithm was applied for digital radiographic images using Bit Slice Processor. In this algorithm, signal enhancing filtering and signal suppressing filtering were performed on the given digital chest image, respectively. Then we grit the dirt- frrence image from these filtered images, and hi-level island images were obtained by applying various threshold values. From the island images, we decided the suspicious nodules using size and circularity test, and marked them to alert radiologists. The performance of the atgorithm was analyzed with respect to the size, contrast and position of digitally synthesized nodules. This method presented 45.8% of true positive ratio for the nodules of lOw in diameter with 12-16 pixel value differnces.

• Radiation Oncology Journal
• /
• 제22권1호
• /
• pp.55-63
• /
• 2004

목적 : 방사선치료계획용 전산화단층활영(computerized tomography, 이하 CT) 시 주사시간이 종양부피의 재현성에 미치는 영향을 살펴보고자 하였다. 대상 및 방법 : 인공호흡기의 환기작용에 의해 세로방향의 왕복 운동을 하는 N자형 모형을 제작하였고, 호흡주기를 T로 하였을 때 주사시간을 상대적으로 각각 0.33, 0.50, 0.67, 0.75, 1.00, 1.33, 그리고 1.53 T로 설정하고 CT촬영을 시행하였다. 또한 2002년 3월 19일부터 2002년 5월 21일까지 서울아산병원 방사선종양학과에서 정위방사선수술(stereotactic radiosurgery)을 시행받은 비소세포성 폐암 환자 3명을 대상으로, 주사시간이 빠른 CT (LightSpeed, GE Medical Systems, 주사시간 0.8초)와 느린 CT (IQ Premier, Picker, 주사시간 2.0초)를 각각 $1\~4$회 시행하였다. 각각의 slice에서, N자형 모형의 왕복 운동이 CT 영상에 반영된 좌우방향 선분의 길이를 측정하여 치료계획용 CT 촬영이 모형의 움직임을 반영하는 정도를 측정하였다. 환자를 대상으로 한 연구에서는 정위적 체부 고정틀을 이용하여, 빠른 CT 및 느린 CT의 종양을 하나의 CT 영상조합에서 재구성하여 종양의 체적과 장경을 측정하여 비교하였다. 결과 : 모형실험에서 주사시간에 비례하여, 세로방향 운동을 CT 촬영에서 반영하는 정도는 증가하였으며 1.00 T 이상에서는 일정한 양상을 보였다 주사시간 1.00 T 이상에서 얻어진 결과를 기준으로, 1.00 T 미만의 주사시간을 가지는 CT촬영에서 모형운동을 반영하지 못하는 비율이 각각 $0.33\;T:\;30\%,\;0.50\;T:\;27\%,\;0.67\;T:\;20\%,\;0.75\;T:\;7.0\%$로 측정되었다. 또한 투시검사로 측정한 종양의 세로 방향 움직임이 각각 3 mm, 5 mm, 10 mm이었던 각각의 환자에서 느린 CT에서 얻어진 종양의 세로방향 장경이 빠른 CT에 비해 $5.3\%\;17\%,\;23\%$ 증가하였다. 결론 : 주사시간을 환자의 호흡주기 이상으로 하는 경우 setup margin만을 고려하여 계획용표적체적(planning target volume, PTV)을 정의할 수 있으므로 정상 폐조직에 조사되는 방사선량을 줄여 치료효율을 향상시킬 수 있을 것으로 생각된다.

• The Korean Journal of Physiology and Pharmacology
• /
• 제16권1호
• /
• pp.59-64
• /
• 2012

Hypoxic pulmonary vasoconstriction (HPV) is physiologically important response for preventing mismatching between ventilation and perfusion in lungs. The HPV of isolated pulmonary arteries (HPV-PA) usually require a partial pretone by thromboxane agonist (U46619). Because the HPV of ventilated/perfused lungs (HPV-lung) can be triggered without pretone conditioning, we suspected that a putative tissue factor might be responsible for the pretone of HPV. Here we investigated whether HPV can be also observed in precision-cut lung slices (PCLS) from rats. The HPV in PCLS also required partial contraction by U46619. In addition, $K^+$ channel blockers (4AP and TEA) required U46619-pretone to induce significant contraction of PA in PCLS. In contrast, the airways in PCLS showed reversible contraction in response to the $K^+$ channel blockers without pretone conditioning. Also, the airways showed no hypoxic constriction but a relaxation under the partial pretone by U46619. The airways in PCLS showed reliable, concentration-dependent contraction by metacholine ($EC_{50}$, ~210 nM). In summary, the HPV in PCLS is more similar to isolated PA than V/P lungs. The metacholineinduced constriction of bronchioles suggested that the PLCS might be also useful for studying airway physiology in situ.

• Korean Journal of Radiology
• /
• 제25권9호
• /
• pp.833-842
• /
• 2024

Objective: To assess the effect of a new lung enhancement filter combined with deep learning image reconstruction (DLIR) algorithm on image quality and ground-glass nodule (GGN) sharpness compared to hybrid iterative reconstruction or DLIR alone. Materials and Methods: Five artificial spherical GGNs with various densities (-250, -350, -450, -550, and -630 Hounsfield units) and 10 mm in diameter were placed in a thorax anthropomorphic phantom. Four scans at four different radiation dose levels were performed using a 256-slice CT (Revolution Apex CT, GE Healthcare). Each scan was reconstructed using three different reconstruction algorithms: adaptive statistical iterative reconstruction-V at a level of 50% (AR50), Truefidelity (TF), which is a DLIR method, and TF with a lung enhancement filter (TF + Lu). Thus, 12 sets of reconstructed images were obtained and analyzed. Image noise, signal-to-noise ratio, and contrast-to-noise ratio were compared among the three reconstruction algorithms. Nodule sharpness was compared among the three reconstruction algorithms using the full-width at half-maximum value. Furthermore, subjective image quality analysis was performed. Results: AR50 demonstrated the highest level of noise, which was decreased by using TF + Lu and TF alone (P = 0.001). TF + Lu significantly improved nodule sharpness at all radiation doses compared to TF alone (P = 0.001). The nodule sharpness of TF + Lu was similar to that of AR50. Using TF alone resulted in the lowest nodule sharpness. Conclusion: Adding a lung enhancement filter to DLIR (TF + Lu) significantly improved the nodule sharpness compared to DLIR alone (TF). TF + Lu can be an effective reconstruction technique to enhance image quality and GGN evaluation in ultralow-dose chest CT scans.

• 한국식생활문화학회지
• /
• 제9권5호
• /
• pp.499-507
• /
• 1995

The purpose of this study was to survey the various kinds of cooked beef products focusing on Sukyuk (boiled beef) and Pyunyuk (boiled beef slice) recorded on the historical literatures written from 1670 to 1945. Sukyuk and Pyunyuk were recorded 45 times in the references and could be classified into 11 groups based on major ingredients such as fresh meat, tough meat, rotten meat, tail, head, lung, cup of breast, testicles, pancreas, spleen and tung. Twenty two cooking methods were described on the records. Sukyuk and Pyunyuk based on fresh meat were described the most frequently. Pyunyuk based on head was described late in 18th centuries, but its recipe could not be found in any records and that of internal organs and tung were presented late in 18th centuries and early in 19th centuries, respectively. The major ingredients of Sukyuk and Pyunyuk were lean meat, tail, head, lung, cup of breast, testicles, pancreas, spleen and tongue Mulberry seed, fragment of roof tile and other sub-ingredients were used for softening or deodorizing the off flavor of the products.

• 대한방사선치료학회지
• /
• 제27권2호
• /
• pp.131-143
• /
• 2015

목 적 : MRI를 기반으로 한 CT fusion plan과 MRI only based plan을 비교함으로써 MRI only plan의 유용성을 평가하고, 나아가 Simulation과 치료계획을 포함한 방사선치료의 전 과정을 CT 촬영 없이 MRI 영상으로 구성하여 실시간 MR-IGRT를 구현하고자 한다. 대상 및 방법 : 본원에서 사용중인 BBB CT (Brilliance Big Bore CT, 16slice, Philips)와 Viewray MRIdian System (Viewray, USA)으로 환자 11명(Head and Neck 1명, Breast 5명, Lung 1명, Liver 3명, Prostate 1명)을 대상으로 CT & MR simulation 및 치료계획을 세웠다. Head and Neck, Breast, Prostate환자는 자유호흡(Free Breathing) 상태로, Lung과 Liver 환자는 흡기 호흡중지(Inhalation Breathing Holdng)상태로 Simulation을 진행하였다. Viewray의 Bore 크기 및 Coil 위치를 고려하여 환자 자세 및 고정기구를 동일한 조건을 유지하여 CT simulation을 시행하였다. Viewray MRIdian 시스템을 이용하여 MR 영상을 기반으로 한 CT fusion plan과 CT fusion 없이 [ICRU 46]에서 권고하는 폐, 공기, 뼈의 전자밀도를 입력한 동일한 조건의 MR only plan을 세웠다. Head and Neck, Breast, Prostate는 IMRT, Lung과 Liver 는 Gating치료계획을 세웠고, 치료계획의 평가는 PTV의 균질성 지표 (Homogeneity Index, HI)와 일치성 지표 (Conformity Index, CI), 그리고 각 PTV와 OAR의 DVH와 선량차이를 각각 비교하였다. 결 과 : 두 치료계획간 치료부위별 PTV에 대한 HI 값의 차이는 Head and Neck, Breast, Lung, Liver, Prostate 부위별로 각각 0.089, 0.26, 0.67, 0.2, 0.4%로 나타났으며, CI 값의 차이는 부위별로 각각 0.043, 0.84, 0.68, 0.46, 0.3%로 두가지 평가 값 모두 Head and Neck 부위가 가장 작은 차이를 나타냈다. PTV에 대한 평균선량 차이는 치료부위별로 각각 0.07, 0.29, 0.18, 0.3, 0.18 Gy로 나타났다. 이를 백분율로 나타냈을 때 0.06, 0.7, 0.29, 0.69, 0.44%으로 모두 1% 이하의 차이를 보였다. 두경부암의 각 OAR은 전체적으로 0.01~0.12 Gy의 평균선량 차이를 보였으며, 유방암은 0.04~0.06 Gy, 폐암에서는 0.01~0.21 Gy, 간암은 0.06~0.27 Gy, 전립선암은 0.02~0.23 Gy의 평균선량 차이를 나타냈다. 결 론 : MR 영상을 이용한 치료계획은 연부조직에서 탁월한 대조도를 나타낼 뿐만 아니라 CT fusion한 MR 치료계획과 비교했을 때 PTV의 HI, CI, 선량차이 모두 1%미만의 차이를 보였으며, OAR의 경우 비균질한 조직이 많은 부위일수록 최대 0.89 Gy 선량차이를 보였다. 이결과를 토대로 두경부암, 부분적 유방암이나 전립선암등 비균질도 차이가 적은 부위에는 CT촬영 없이 MR 영상만을 이용한 방사선치료의 가능성을 확인 할 수 있었다.