• Korean Journal of Radiology
• /
• v.22 no.3
• /
• 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.

• Tuberculosis and Respiratory Diseases
• /
• v.67 no.1
• /
• pp.37-41
• /
• 2009

Lymphocytic interstitial pneumonia (LIP) is a rare disorder characterized by a diffuse infiltration of the alveolar space, interstitium by lymphocytes, plasma cells, and reticuloendothelial cells. Although its etiology is unknown, LIP has been associated with autoimmune disorders and with viral infections. Because it's clinical and radiographic features are nonspecific, a confirmatory diagnosis is performed by open lung biopsy. A 59-year-old female presented dry cough, which had been present for 1 month. On initial findings of multifocal consolidation at the right middle lobe on both lower lobes in chest radiography, the first diagnosis of cryptogenic organizing pneumonia was suggested. On open lung biopsy, LIP was diagnosed. The patient had no autoimmune disease, viral infection or monoclonal gammopathy. After 3 months of corticosteroid treatment, the patient experienced improved symptoms, reduced abnormalities on chest radiography, and improved pulmonary function testing.

• Journal of Veterinary Clinics
• /
• v.24 no.1
• /
• pp.46-50
• /
• 2007

A 7-year-old female, Jindo dog was referred to the Veterinary Teaching Hospital with mild dyspnea and anorexia due to a gunshot trauma. The dog was wounded in the thoracic region 3 days ago. Plain radiographs showed the left 8th rib fracture, interstitial pattern in the left caudal lung field and pleural effusion. Abdominal radiographs showed the lead bullet. Computed tomographs(CT) showed the size of pulmonary contusion, laceration, lung parenchymal injuries, hemothorax and perforation of abdominal wall. The therapeutic plan was based on abnormalities seen on CT scans but not clearly seen in survey radiographs. Thoracic CT significantly provides even more informations compared with the corresponding radiographs in thoracic gunshot trauma. Although thoracic survey radiographs are useful as a screening tool, CT is highly sensitive in detecting thoracic injuries after thoracic trauma and is superior to routine thoracic survey radiographs in visualizing lung contusion, pneumothorax and hemothorax. Therefore, we recommend CT in the initial diagnostic work-up of patients with thoracic injuries and with suspected chest trauma because early and exact diagnosis of all thoracic injuries along with sufficient therapeutic consequences may reduce complications.

• Journal of Preventive Medicine and Public Health
• /
• v.29 no.4 s.55
• /
• pp.853-861
• /
• 1996

Pneumoconiosis is one of the major problem in the field of occupational health at Korea. Therefore, the efficient diagnosis of pneumoconiosis is a hot issue on the occupational health program. The author executed this study to estimate the diagnostic value of high resolution computed tomography(HRCT) compared with chest radiography for screening of welder's lung. HRCT was introduced very recently for the diagnosis of pneumoconiosis, however, the diagnostic value for screening of welder's lung - principally nonfibrogenic and reversible - has not been evaluated. The subjects were fifty cases of welder's lung or suspected cases who had been collected between 1989 and 1994 from one shipyard and continuously followed-up on the basis of in-plant periodic health check program. We applied both chest radiography and HRCT on the same subjects from May 1 to 30, 1996. The images were evaluated by two careered radiologists independently. The findings of chest radiography were classified into four category by ILO classification, and the findings of HRCT according to the criteria of Bergin et al. The concordance between two radiologists expressed with Kendall's tau-b was 0.72 by chest radiography and 0.44 by HRCT- that is, interobserver variation of HRCT was bigger than that of chest radiography. The concordance between the two different methods was highly variable as 0.44 by radiologist A and 0.06 by radiologist B - that is, interobserver variation was very big. However, HRCT looked more detectable for the minor parenchymal change. These findings suggested that it is not appropriate to use HRCT routinely for screening of welder's lung due to lack of diagnostic criteria, and feasibility, acceptability and economic aspects. Nevertheless, HRCT might be recommendable in the case of equivocal parenchymal features on the chest radiography, unexplained respiratory symptoms, and/or lung function abnormalities suggestive of interstitial fibrosis.

• Tuberculosis and Respiratory Diseases
• /
• v.45 no.1
• /
• pp.5-19
• /
• 1998

Introduction: Direct histologic and bacteriologic examination of a representative specimen of lung tissue is the only certain method of providing an accurate diagnosis in various pulmonary diseases including diffuse pulmonary diseases. The purpose of national survey was to define the indication, incidence, effectiveness, safety and complication of open and thoracoscopic lung biopsy in korea. Methods: A multicenter registry of 37 university or general hospitals equipped more than 400 patient's bed were retrospectively collected and analyzed for 3 years from the January 1994 to December 1996 using the same registry protocol. Results: 1) There were 511 cases from the 37 hospitals during 3 years. The mean age was 50.2 years(${\pm}15.1$ years) and men was more prevalent than women(54.9% vs 45.9%). 2) The open lung biopsy was performed in 313 cases(62%) and thoracoscopic lung biopsy was performed in 192 cases(38%). The incidence of lung biopsy was more higher in diffuse lung disease(305 cases, 59.7%) than in localized lung disease(206 cases, 40.3%) 3) The duration after abnormalities was found in chest X-ray until lung biopsy was 82.4 days(open lung biopsy: 72.8 days, thoracoscopic lung biopsy: 99.4 days). The bronchoscopy was performed in 272 cases(53.2%), bronchoalveolar lavage was performed in 123 cases(24.1%) and percutaneous lung biopsy was performed in 72 cases(14.1%) before open or thoracoscopic lung biopsy. 4) There were 230 cases(45.0%) of interstitial lung disease, 133 cases(26.0%) of thoracic malignancies, 118 cases(23.1%) of infectious lung disease including tuberculosis and 30 cases (5.9 %) of other lung diseases including congenital anomalies. No significant differences were noted in diagnostic rate and disease characteristics between open lung biopsy and thoracoscopic lung biopsy. 5) The final diagnosis through an open or thoracoscopic lung biopsy was as same as the presumptive diagnosis before the biopsy in 302 cases(59.2%). The identical diagnostic rate was 66.5% in interstitial lung diseases, 58.7% in thoracic malignancies, 32.7% in lung infections, 55.1 % in pulmonary tuberculosis, 62.5% in other lung diseases including congenital anomalies. 6) One days after lung biopsy, $PaCO_2$ was increased from the prebiopsy level of $38.9{\pm}5.8mmHg$ to the $40.2{\pm}7.1mmHg$(P<0.05) and $PaO_2/FiO_2$ was decreased from the prebiopsy level of $380.3{\pm}109.3mmHg$ to the $339.2{\pm}138.2mmHg$(P=0.01). 7) There was a 10.1 % of complication after lung biopsy. The complication rate in open lung biopsy was much higher than in thoracoscopic lung biopsy(12.4% vs 5.8%, P<0.05). The incidence of complication was pneumothorax(23 cases, 4.6%), hemothorax(7 cases, 1.4%), death(6 cases, 1.2%) and others(15 cases, 2.9%). 8) The 5 cases of death due to lung biopsy were associated with open lung biopsy and one fatal case did not describe the method of lung biopsy. The underlying disease was 3 cases of thoracic malignancies(2 cases of bronchoalveolar cell cancer and one malignant mesothelioma), 2 cases of metastatic lung cancer, and one interstitial lung disease. The duration between open lung biopsy and death was $15.5{\pm}9.9$ days. 9) Despite the lung biopsy, 19 cases (3.7%) could not diagnosed. These findings were caused by biopsy was taken other than target lesion(5 cases), too small size to interpretate(3 cases), pathologic inability(11 cases). 10) The contribution of open or thoracoscopic lung biopsy to the final diagnosis was defininitely helpful(334 cases, 66.5%), moderately helpful(140 cases, 27.9%), not helpful or impossible to judge(28 cases, 5.6%). Overall, open or thoracoscopic lung biopsy were helpful to diagnose the lung lesion in 94.4 % of total cases. Conclusions: The open or thoracoscopic lung biopsy were relatively safe and reliable diagnostic method of lung lesion which could not diagnosed by other diagnostic approaches such as bronchoscopy. We recommend the thoracoscopic lung biopsy when the patients were in critical condition because the thoracoscopic biopsy was more safe and have equal diagnostic results compared with the open lung biopsy.

• Journal of Veterinary Clinics
• /
• v.40 no.4
• /
• pp.308-313
• /
• 2023

A 1-year-old castrated male Korean Shorthair cat presented with dyspnea, anorexia, lethargy, and seizures. Physical examination revealed salivation, right forelimb hemiparesis, and rapid breathing. No abnormalities were detected on auscultation. Laboratory findings revealed increased levels of bilirubin, aspartate aminotransferase (AST), globulin, glucose, and a decreased albumin-to-globulin (A:G) ratio. Both N-terminal pro-B-type natriuretic peptide (NT-proBNP) and feline serum amyloid A (fSAA) levels were significantly elevated. Thoracic radiography revealed mild cardiomegaly and diffuse increased interstitial infiltration with soft tissue opacity in the periphery of the right caudal pleural space. Echocardiography and lung ultrasonography were performed to investigate the cause of mild cardiomegaly and soft tissue opacity in the pleural space. Echocardiography revealed a mild amount of echogenic pericardial effusion, and lung ultrasonography showed an echogenic soft tissue mass with no blood signal in the right caudal pleural space, suggestive of a granulomatous lesion. After obtaining 5 mL of pericardial fluid through pericardiocentesis, cytology of the pericardial effusion sample revealed marked neutrophils and macrophages with no bacteria. IDEXX feline infectious peritonitis (FIP) virus real-time reverse transcriptase polymerase chain reaction (RT-PCR) confirmed the presence of the FIP virus biotype in the sample. This case presents a rarely reported atypical mixed form of FIP in a cat diagnosed ante-mortem using pericardial effusion analysis. In this case, ultrasound examination played a crucial role in the definitive diagnosis of FIP by PCR biotyping through pericardiocentesis. Ultrasonography can be highly beneficial in guiding the diagnosis and evaluation of cats with suspected FIP.

• Korean Journal of Radiology
• /
• v.24 no.8
• /
• pp.807-820
• /
• 2023

Objective: To assess whether computed tomography (CT) conversion across different scan parameters and manufacturers using a routable generative adversarial network (RouteGAN) can improve the accuracy and variability in quantifying interstitial lung disease (ILD) using a deep learning-based automated software. Materials and Methods: This study included patients with ILD who underwent thin-section CT. Unmatched CT images obtained using scanners from four manufacturers (vendors A-D), standard- or low-radiation doses, and sharp or medium kernels were classified into groups 1-7 according to acquisition conditions. CT images in groups 2-7 were converted into the target CT style (Group 1: vendor A, standard dose, and sharp kernel) using a RouteGAN. ILD was quantified on original and converted CT images using a deep learning-based software (Aview, Coreline Soft). The accuracy of quantification was analyzed using the dice similarity coefficient (DSC) and pixel-wise overlap accuracy metrics against manual quantification by a radiologist. Five radiologists evaluated quantification accuracy using a 10-point visual scoring system. Results: Three hundred and fifty CT slices from 150 patients (mean age: 67.6 ± 10.7 years; 56 females) were included. The overlap accuracies for quantifying total abnormalities in groups 2-7 improved after CT conversion (original vs. converted: 0.63 vs. 0.68 for DSC, 0.66 vs. 0.70 for pixel-wise recall, and 0.68 vs. 0.73 for pixel-wise precision; P < 0.002 for all). The DSCs of fibrosis score, honeycombing, and reticulation significantly increased after CT conversion (0.32 vs. 0.64, 0.19 vs. 0.47, and 0.23 vs. 0.54, P < 0.002 for all), whereas those of ground-glass opacity, consolidation, and emphysema did not change significantly or decreased slightly. The radiologists' scores were significantly higher (P < 0.001) and less variable on converted CT. Conclusion: CT conversion using a RouteGAN can improve the accuracy and variability of CT images obtained using different scan parameters and manufacturers in deep learning-based quantification of ILD.

• Journal of Veterinary Clinics
• /
• v.29 no.1
• /
• pp.78-81
• /
• 2012

A 5-year-old male Pointer dog weighing 24 kg was referred because of a chronic cough. The dog was routinely vaccinated and treated monthly with medications to prevent heartworm and endoparasites. Blood analyses revealed no abnormalities. Chest radiographs showed interstitial and bronchial patterns in the caudal lung fields. Bronchoscopic imaging identified a collapse of the left bronchus with a mucopurulent discharge, indicating the onset of bronchomalacia. Computed tomography also showed bronchial collapse and a tapering angle in the left caudal bronchus. This case study describes the diagnosis of bronchomalacia in a Pointer dog with a chronic cough, and suggests that bronchoscopic examination is a valuable tool for identifying of the onset of bronchomalacia in dogs presenting with an uncontrolled cough.

• Tuberculosis and Respiratory Diseases
• /
• v.51 no.4
• /
• pp.346-353
• /
• 2001

Background : Pulmonary Langerhans cell histiocytosis forms part of a spectrum of diseases that are characterized by a monoclonal proliferation and infiltration of organs by Langerhans cells. Several organ systems may be involved in Langerhans cell histiocytosis, including the lungs, bone, skin, pituitary gland, liver, lymph nodes and thyroid. Pulmonary histiocytosis X represents 2.8% of interstitial lung disease. Here we present the clinical, radiological, therapeutic aspects of pulmonary histiocytosis X. Method : Fourteen cases of biopsy-proven pulmonary histiocytosis X patients who were diagnosed in Seoul National University Hospital during the period from January 1990 to December 1998 were analyzed retrospectively. Result : There were 12 men and 2 women in this study. The initial presenting symptoms were dyspnea, cough, chest pain, which was associated with the pneumothorax, and chest radiography abnormalities. Only 8 patients (57%) were smokers. There were 5 patients with extra-pulmonary histiocytosis (pituitary, bone, skin). Eight patients had received the chemotherapy. There were no mortalities and only one patient experienced an aggravation of symptom during the follow-up period. Conclusion : In contrast to previous reports from other countries, the patients with pulmonary histiocytosis X in this study presented with several different clinical characteristics, such as a male predominance, relatively low smoker's rate, and a better prognosis.

• Tuberculosis and Respiratory Diseases
• /
• v.38 no.2
• /
• pp.155-163
• /
• 1991

Bronchoalveolar lavage had been done as the treatment of some diseases such as alveolar proteinsois, bronchiectasis, and severe asthma to remove excessive secretion or mucus. But in the recent decade it has been done as a diagnostic method and a tool to understand and evaluate the pathophysiology of diffuse interstitial lung diseases such as sarcoidosis, pneumoconiosis and hypersensitivity pneumonitis. To analyse the bronchoalveolar fluid, it might be useful to have a standard reference (especially cell counts and differetial count of the cells from bronchoalveolar lavage fluid) of normal person. But it is difficult to study the normal volunteers. We investgated the bronchoalveolar lavage fluid of 48 patients (28 nonsmokers, 20 smokers) who visited Severance Hospital because of minor pulmonary symptoms such as cough and sputum. They did neither complain of dyspnea nor cyanosis, and had normal or unilateral minor lesion on physical examination and chest X-ray. We analysed the recovery rate, viability, total cell count and differential count of the cells in fluid obtained by bronchoalveolar lavage. The following results were obtained: 1) Age ranged from 17 to 72 years-old with the mean age of 36.7; there was no difference of age between the nonsmoker and the smoker gorup. Male to female ratio was 2.43:1 for total group, 1.15:1 for nonsmokers, and 19:1 for smokers. 2) The diagnoses of the patients were undetermined in 41.9%, healed pulmonary tuberculosis in 37.5%, laryngitis or pharyngitis in 10.4% and others in 10.4%. 3) Total cell number of the recovered fluid by bronchoalveolar lavage was significantly higher in male[$9.6{\pm}6.2({\times}10^6)$] than in female[$5.1{\pm}3.0({\times}10^6)$](p<0.05), and there was no significant difference in the total cell number between the smokers and nonsmokers [$9.3{\pm}5.8({\times}10^6)$ vs $7.5{\pm}5.8({\times}10^6)$]. 4) The differential count of the cells from bronchoalveolar lavage fluid had no difference between the nonsmokers and the smokers. 5) There was no correlation between the total cell count and smoking or age. 6) In the smoker group, there was no correlation between the amount of smoking and the total cell count of the bronchoalveolar fluid. In conclusion, it should be careful to regard the patients with symptoms or minor radiologic abnormalities as a control group in bronchoalveolar lavage study and further study of cell analysis in bronchoalveolar lavage will be needed between smoker and nonsmoker in the male and female healthy people.