Content-Based Image Retrieval of Chest CT with Convolutional Neural Network for Diffuse Interstitial Lung Disease: Performance Assessment in Three Major Idiopathic Interstitial Pneumonias

Objective@#To assess the performance of content-based image retrieval (CBIR) of chest CT for diffuse interstitial lung disease (DILD). @*Materials and Methods@#The database was comprised by 246 pairs of chest CTs (initial and follow-up CTs within two years) from 246 patients with usual interstitial pneumonia (UIP, n = 100), nonspecific interstitial pneumonia (NSIP, n = 101), and cryptogenic organic pneumonia (COP, n = 45). Sixty cases (30-UIP, 20-NSIP, and 10-COP) were selected as the queries. The CBIR retrieved five similar CTs as a query from the database by comparing six image patterns (honeycombing, reticular opacity, emphysema, ground-glass opacity, consolidation and normal lung) of DILD, which were automatically quantified and classified by a convolutional neural network. We assessed the rates of retrieving the same pairs of query CTs, and the number of CTs with the same disease class as query CTs in top 1–5 retrievals. Chest radiologists evaluated the similarity between retrieved CTs and queries using a 5-scale grading system (5-almost identical; 4-same disease; 3-likelihood of same disease is half; 2-likely different; and 1-different disease). @*Results@#The rate of retrieving the same pairs of query CTs in top 1 retrieval was 61.7% (37/60) and in top 1–5 retrievals was 81.7% (49/60). The CBIR retrieved the same pairs of query CTs more in UIP compared to NSIP and COP (p = 0.008 and 0.002).On average, it retrieved 4.17 of five similar CTs from the same disease class. Radiologists rated 71.3% to 73.0% of the retrieved CTs with a similarity score of 4 or 5. @*Conclusion@#The proposed CBIR system showed good performance for retrieving chest CTs showing similar patterns for DILD.

A Rare Case of Primary Aortoduodenal Fistula

Aortoduodenal fistula is a rare but life-threatening condition that can cause gastrointestinal bleeding. Due to its rarity, it is often overlooked as a cause of gastrointestinal blood loss. Notably, the mortality rate of aortoduodenal fistula is nearly 100% in undiagnosed and untreated cases. We report a case of aortoduodenal fistula, which resulted in the patient’s death. This report highlights the importance of considering even extremely rare causes of gastrointestinal bleeding in the differential diagnosis in patients with such a presentation.

Convolutional Neural Network Technology in Endoscopic Imaging: Artificial Intelligence for Endoscopy

Recently, significant improvements have been made in artificial intelligence. The artificial neural network was introduced in the 1950s. However, because of the low computing power and insufficient datasets available at that time, artificial neural networks suffered from overfitting and vanishing gradient problems for training deep networks. This concept has become more promising owing to the enhanced big data processing capability, improvement in computing power with parallel processing units, and new algorithms for deep neural networks, which are becoming increasingly successful and attracting interest in many domains, including computer vision, speech recognition, and natural language processing. Recent studies in this technology augur well for medical and healthcare applications, especially in endoscopic imaging. This paper provides perspectives on the history, development, applications, and challenges of deep-learning technology.

A Rare Fatal Bile Peritonitis after Malposition of Endoscopic Ultrasound-Guided 5-Fr Naso-Gallbladder Drainage

Endoscopic ultrasound (EUS)-guided gallbladder (GB) drainage has recently emerged as a more feasible treatment than percutaneous transhepatic GB drainage for acute cholecystitis. In EUS-guided cholecystostomies in patients with distended GBs without pericholecystic inflammation or prominent wall thickening, a needle puncture with tract dilatation is often difficult. Guidewires may slip during the insertion of thin and flexible drainage catheters, which can also cause the body portion of the catheter to be unexpectedly situated and prolonged between the GB and intestines because the non-inflamed distended GB is fluctuant. Upon fluoroscopic examination during the procedure, the position of the abnormally coiled catheter may appear to be correct in patients with a distended stomach. We experienced such an adverse event with fatal bile peritonitis in a patient with GB distension suggestive of malignant bile duct stricture. Fatal bile peritonitis then occurred. Therefore, the endoscopist should confirm the indications for cholecystostomy and determine whether a distended GB is a secondary change or acute cholecystitis.

Data Augmentation Techniques for Deep Learning-Based Medical Image Analyses

Medical image analyses have been widely used to differentiate normal and abnormal cases, detect lesions, segment organs, etc. Recently, owing to many breakthroughs in artificial intelligence techniques, medical image analyses based on deep learning have been actively studied. However, sufficient medical data are difficult to obtain, and data imbalance between classes hinder the improvement of deep learning performance. To resolve these issues, various studies have been performed, and data augmentation has been found to be a solution. In this review, we introduce data augmentation techniques, including image processing, such as rotation, shift, and intensity variation methods, generative adversarial network-based method, and image property mixing methods. Subsequently, we examine various deep learning studies based on data augmentation techniques. Finally, we discuss the necessity and future directions of data augmentation.

Data Augmentation Techniques for Deep Learning-Based Medical Image Analyses

Medical image analyses have been widely used to differentiate normal and abnormal cases, detect lesions, segment organs, etc. Recently, owing to many breakthroughs in artificial intelligence techniques, medical image analyses based on deep learning have been actively studied. However, sufficient medical data are difficult to obtain, and data imbalance between classes hinder the improvement of deep learning performance. To resolve these issues, various studies have been performed, and data augmentation has been found to be a solution. In this review, we introduce data augmentation techniques, including image processing, such as rotation, shift, and intensity variation methods, generative adversarial network-based method, and image property mixing methods. Subsequently, we examine various deep learning studies based on data augmentation techniques. Finally, we discuss the necessity and future directions of data augmentation.

Risk Factors for Asymptomatic Colon Diverticulosis / 대한소화기병학회잡지

BACKGROUND/AIMS: The etiology of colon diverticulosis is related to a range of genetic, biological, and environmental factors, but the risk factors for asymptomatic diverticulosis of the colon are unclear. This study examined the risk factors for asymptomatic colon diverticulosis. METHODS: This retrospective study included examinees who underwent a colonoscopy for screening at the health check-up center of SAM Hospital between January 2016 and December 2016. The examinees with colon diverticulosis found by colonoscopy were compared with those without diverticulosis. The comparison factors were age, gender, alcohol consumption, smoking status, medical history, lipid profile, body mass index, visceral fat area, waist-hip ratio, and severity of a fatty liver. RESULTS: This study included 937 examinees and the overall prevalence of diverticulosis was 8.1% (76/937). Fatty liver was found in 69.7% (53/76) in cases of colon diverticulosis and 50.3% (433/861) in the control group (p=0.001). The average waist-hip ratio was 0.92±0.051 in colon diverticulosis and 0.90±0.052 in the control group (p=0.052). Multivariate analysis revealed the waist-hip ratio (OR=1.035, 95% CI 1.000–1.070, p=0.043), moderate fatty liver (OR=2.238, 95% CI 1.026–4.882, p=0.043), and severe fatty liver (OR=5.519, 95% CI 1.236–21.803, p=0.025) to be associated with an increased risk of asymptomatic colon diverticulosis. CONCLUSIONS: The waist-hip ratio, moderate fatty liver, and severe fatty liver are risk factors for asymptomatic colon diverticulosis. Central obesity, which can be estimated by the waist-hip ratio, and fatty liver might affect the pathogenesis of asymptomatic colon diverticulosis.

Risk Factors for Asymptomatic Colon Diverticulosis / 대한소화기학회지

BACKGROUND/AIMS: The etiology of colon diverticulosis is related to a range of genetic, biological, and environmental factors, but the risk factors for asymptomatic diverticulosis of the colon are unclear. This study examined the risk factors for asymptomatic colon diverticulosis.METHODS: This retrospective study included examinees who underwent a colonoscopy for screening at the health check-up center of SAM Hospital between January 2016 and December 2016. The examinees with colon diverticulosis found by colonoscopy were compared with those without diverticulosis. The comparison factors were age, gender, alcohol consumption, smoking status, medical history, lipid profile, body mass index, visceral fat area, waist-hip ratio, and severity of a fatty liver.RESULTS: This study included 937 examinees and the overall prevalence of diverticulosis was 8.1% (76/937). Fatty liver was found in 69.7% (53/76) in cases of colon diverticulosis and 50.3% (433/861) in the control group (p=0.001). The average waist-hip ratio was 0.92±0.051 in colon diverticulosis and 0.90±0.052 in the control group (p=0.052). Multivariate analysis revealed the waist-hip ratio (OR=1.035, 95% CI 1.000–1.070, p=0.043), moderate fatty liver (OR=2.238, 95% CI 1.026–4.882, p=0.043), and severe fatty liver (OR=5.519, 95% CI 1.236–21.803, p=0.025) to be associated with an increased risk of asymptomatic colon diverticulosis.CONCLUSIONS: The waist-hip ratio, moderate fatty liver, and severe fatty liver are risk factors for asymptomatic colon diverticulosis. Central obesity, which can be estimated by the waist-hip ratio, and fatty liver might affect the pathogenesis of asymptomatic colon diverticulosis.

Concomitant or Sequential Therapy as the First-line Therapy for Eradication of Helicobacter pylori Infection in Korea: A Systematic Review and Meta-analysis / 대한소화기학회지

BACKGROUND/AIMS: In Korea, increasing clarithromycin resistance has led to the need for an alternative first-line therapy for the eradication of Helicobacter pylori (H. pylori) infection. Concomitant therapy (CT) and sequential therapy (ST) have been proposed as alternative regimens. The aim of this study was to compare the eradication rate from using CT and ST in Korea. METHODS: A literature review was performed on studies comparing the efficacy of CT and ST in Korea. Data were pooled to obtain the odds ratio (OR) of the eradication rate with 95% confidence intervals (CIs). The eradication rates were considered both on an intention-to-treat (ITT) and a per-protocol (PP) bases. RESULTS: Six studies provided data on 1,897 Korean adult patients. The pooled OR was 1.382 (95% CI: 1.031-1.853, p=0.031) for ITT analysis and 2.114 (95% CI: 1.502-2.974, p < 0.001) for PP analysis. There was no difference in the rate of adverse events and compliances between the two regimens. CONCLUSIONS: The efficacy of CT was superior to ST in both ITT and PP analyses. Therefore, CT could be an excellent alternative regimen for the eradication of H. pylori as a first-line therapy in Korea.

Solitary Jejunal Tuberculosis with Intestinal Obstruction in an Immunocompetent Patient / 대한내과학회지

Intestinal tuberculosis is an infection of the gastrointestinal tract by the Mycobacterium tuberculosis complex. To the best of our knowledge, solitary intestinal tuberculosis accompanied by intestinal obstruction, particularly in the middle of the small intestine, is extremely rare. We report a case of solitary jejunal tuberculosis in a 49-year-old man with no underlying disease. He was admitted a few days after the onset of diffuse abdominal discomfort. Upon evaluation, we initially considered a malignancy of the distal jejunum with ileus due to the presence of a mass. Therefore, he underwent laparoscopic resection of the small bowel. Unexpectedly, the histologic specimen showed a chronic caseating granulomatous lesion with acid-fast bacilli. Ultimately, he was diagnosed with solitary jejunal tuberculosis. He was successfully treated with anti-tuberculosis drugs without any complications.

Development of a Rapid Automated Fluorescent Lateral Flow Immunoassay to Detect Hepatitis B Surface Antigen (HBsAg), Antibody to HBsAg, and Antibody to Hepatitis C

BACKGROUND: Accurate, rapid, and cost-effective screening tests for hepatitis B virus (HBV) and hepatitis C virus (HCV) infection may be useful in laboratories that cannot afford automated chemiluminescent immunoassays (CLIAs). We evaluated the diagnostic performance of a novel rapid automated fluorescent lateral flow immunoassay (LFIA). METHODS: A fluorescent LFIA using a small bench-top fluorescence reader, Automated Fluorescent Immunoassay System (AFIAS; Boditech Med Inc., Chuncheon, Korea), was developed for qualitative detection of hepatitis B surface antigen (HBsAg), antibody to HBsAg (anti-HBs), and antibody to HCV (anti-HCV) within 20 minutes. We compared the diagnostic performance of AFIAS with that of automated CLIAs—Elecsys (Roche Diagnostics GmbH, Penzberg, Germany) and ARCHITECT (Abbott Laboratories, Abbott Park, IL, USA)—using 20 seroconversion panels and 3,500 clinical serum samples. RESULTS: Evaluation with the seroconversion panels demonstrated that AFIAS had adequate sensitivity for HBsAg and anti-HCV detection. From the clinical samples, AFIAS sensitivity and specificity were 99.8% and 99.3% for the HBsAg test, 100.0% and 100.0% for the anti-HBs test, and 98.8% and 99.1% for the anti-HCV test, respectively. Its agreement rates with the Elecsys HBsAg, anti-HBs, and anti-HCV detection assays were 99.4%, 100.0%, and 99.0%, respectively. AFIAS detected all samples with HBsAg genotypes A-F and H and anti-HCV genotypes 1, 1a, 1b, 2a, 2b, 4, and 6. Cross-reactivity with other infections was not observed. CONCLUSIONS: The AFIAS HBsAg, anti-HBs, and anti-HCV tests demonstrated diagnostic performance equivalent to current automated CLIAs. AFIAS could be used for a large-scale HBV or HCV screening in low-resource laboratories or low-to middle-income areas.

Development of Kaposi sarcoma and hemophagocytic lymphohistiocytosis associated with human herpesvirus 8 in a renal transplant recipient

No abstract available.