OBJECTIVES: We propose a multi-feature fusion model based on manually extracted features and deep learning features from endoscopic images for grading rebleeding risk of peptic ulcers. METHODS: Based on the endoscopic appearance of peptic ulcers, color features were extracted to distinguish active bleeding (Forrest I) from non-bleeding ulcers (Forrest II and III). The edge and texture features were used to describe the morphology and appearance of the ulcers in different grades. By integrating deep features extracted from a deep learning network with manually extracted visual features, a multi-feature representation of endoscopic images was created to predict the risk of rebleeding of peptic ulcers. RESULTS: In a dataset consisting of 3573 images from 708 patients with Forrest classification, the proposed multi-feature fusion model achieved an accuracy of 74.94% in the 6-level rebleeding risk classification task, outperforming the experienced physicians who had a classification accuracy of 59.9% (P<0.05). The F1 scores of the model for identifying Forrest Ib, IIa, and III ulcers were 90.16%, 75.44%, and 77.13%, respectively, demonstrating particularly good performance of the model for Forrest Ib ulcers. Compared with the first model for peptic ulcer rebleeding classification, the proposed model had improved F1 scores by 5.8%. In the simplified 3-level risk (high-risk, low-risk, and non-endoscopic treatment) classification task, the model achieved F1 scores of 93.74%, 81.30%, and 73.59%, respectively. CONCLUSIONS The proposed multi-feature fusion model integrating deep features from CNNs with manually extracted visual features effectively improves the accuracy of rebleeding risk classification for peptic ulcers, thus providing an efficient diagnostic tool for clinical assessment of rebleeding risks of peptic ulcers.
Humans ; Deep Learning ; Peptic Ulcer ; Risk Assessment ; Peptic Ulcer Hemorrhage ; Recurrence

Bleeding peptic ulcers remained as one of the commonest causes of hospitalization worldwide. While endoscopic hemostasis serves as primary treatment for bleeding ulcers, rebleeding after endoscopic hemostasis becomes more and more difficult to manage as patients are usually poor surgical candidates with multiple comorbidities. Recent advances in management of bleeding peptic ulcers aimed to further reduce the rate of rebleeding through—(1) identification of high risk patients for rebleeding and mortality; (2) improvement in primary endoscopic hemostasis and; (3) prophylactic angiographic embolization of major arteries. The technique and clinical evidences for these approaches will be reviewed in the current article.
Arteries ; Comorbidity ; Endoscopy ; Hemorrhage ; Hemostasis, Endoscopic ; Hospitalization ; Humans ; Mortality ; Peptic Ulcer ; Ulcer

Non-variceal upper gastrointestinal bleeding (NVUGIB) refers to blood loss from the gastrointestinal tract proximal to the ligament of Treitz due to lesions that are non-variceal in origin. The distinction of the bleeding source as non-variceal is important in numerous aspects, but none more so than endoscopic approaches for successful hemostasis. When a patient presents with acute overt blood loss, NVUGIB is a medical emergency, which requires immediate intervention. There have been major strides in pharmacologic and endoscopic interventions for successful induction and remission of hemostasis in the last two decades. Despite achieving tangible improvements, the burden of the disease and the consequent mortality remain high. To address endoscopic outcomes better, several new technologies have emerged and have been subsequently incorporated to the armamentarium of hemostatic tools. This study aims to provide a succinct review on novel technologies for endoscopic hemostasis.
Emergencies ; Endoscopy ; Gastrointestinal Hemorrhage ; Gastrointestinal Tract ; Hemorrhage ; Hemostasis ; Hemostasis, Endoscopic ; Humans ; Ligaments ; Mortality ; Peptic Ulcer

Among gastrointestinal emergencies, acute upper gastrointestinal bleeding remains a challenging clinical problem owing to significant patient morbidity and costs involved in management. Endoscopic hemostatic therapy is the mainstay of treatment and decreases the incidence of re-bleeding, the need for surgery, morbidity, and mortality. However, in 8%–15% of patients with upper gastrointestinal bleeding, endoscopic hemostatic therapy does not successfully control bleeding. Trans-arterial coil embolization is an effective alternative treatment for endoscopic hemostatic failure; however, this procedure can induce adverse outcomes, such as non-target vessel occlusion, vessel dissection and perforation, and coil migration. Coil migration is rare but causes severe complications, such as re-bleeding and bowel ischemia. However, in most cases, coil migration is local and involves spontaneous healing without serious complications. Here, we report the case of a patient who underwent trans-arterial coil embolization of the gastroduodenal artery with the purpose of controlling massive duodenal bleeding, resulting in a fatal outcome caused by coil migration.
Arteries ; Embolization, Therapeutic ; Emergencies ; Endoscopy ; Fatal Outcome ; Hemorrhage ; Humans ; Incidence ; Ischemia ; Mortality ; Peptic Ulcer Hemorrhage

No abstract available.
Helicobacter pylori ; Helicobacter ; Hemorrhage ; Peptic Ulcer

BACKGROUND/AIMS: The association between Helicobacter pylori infection and nonsteroidal anti-inflammatory drugs (NSAIDs) or low-dose aspirin therapy as a risk factor for peptic ulcer bleeding (PUB) remains unclear. This study investigated the risk of PUB associated with H. pylori infection and NSAID or low-dose aspirin therapy in patients with PUD. MATERIALS AND METHODS: This case-control study investigated 340 patients with PUB between 2012 and 2016. The control group comprised age and sex-matched patients with endoscopically documented non-bleeding ulcers. Using logistic regression analysis, the adjusted odds ratio (AOR) was calculated for the risk of PUB. RESULTS: Of the patients investigated, 57.9% in the study group and 51.8% in the control group were diagnosed with H. pylori infection (P=0.106). Logistic regression analysis showed synergistic interaction between H. pylori infection and low-dose aspirin therapy. Multivariate analysis showed that low-dose aspirin (AOR 3.92, P < 0.001), NSAIDs (AOR 2.98, P=0.001), warfarin (AOR 14.57, P=0.011), gastric ulcer (compared with duodenal ulcer) (AOR 1.65, P=0.01), and smoking (AOR 1.97, P=0.004) increased the risk of PUB compared with the risk of PUD. CONCLUSIONS: Both NSAIDs and aspirin are independent risk factors for bleeding in patients with PUD. Additionally, low-dose aspirin therapy concomitant with H. pylori infection produced a synergistic effect. Therefore, H. pylori eradication may be crucial in aspirin users. Moreover, a proton pump inhibitor should be prescribed in patients with a history of bleeding ulcers who need long-term NSAID treatment.
Anti-Inflammatory Agents, Non-Steroidal ; Aspirin ; Case-Control Studies ; Helicobacter pylori ; Helicobacter ; Hemorrhage ; Humans ; Logistic Models ; Multivariate Analysis ; Odds Ratio ; Peptic Ulcer ; Proton Pumps ; Risk Factors ; Smoke ; Smoking ; Stomach Ulcer ; Ulcer ; Warfarin

Intramural hematoma of the duodenum is a relatively unusual complication associated with the endoscopic treatment of bleeding peptic ulcers. Intramural hematomas are typically resolved spontaneously with conservative treatment alone. We report a case of an intramural duodenal hematoma following endoscopic hemostasis with epinephrine injection therapy, which was associated with transient obstructive jaundice in a patient undergoing hemodialysis. The patient developed biliary sepsis due to obstruction of the common bile duct secondary to the huge hematoma. He was treated with fluoroscopy-guided drainage catheter insertion, which spontaneously resolved the biliary sepsis through conservative treatment in 6 weeks. Fluoroscopy-guided drainage may impact the treatment of intramural hematomas that involve life-threatening complications.
Catheters ; Cholestasis ; Common Bile Duct ; Drainage ; Duodenal Ulcer ; Duodenum ; Epinephrine ; Hematoma* ; Hemorrhage ; Hemostasis, Endoscopic* ; Humans ; Jaundice, Obstructive* ; Peptic Ulcer ; Renal Dialysis ; Sepsis

PURPOSE: This study was performed to assess the clinical usefulness of transabdominal ultrasonography (TUS) in detecting peptic ulcer disease (PUD) in children. METHODS: Twenty-four patients (19 boys, 5 girls; mean age, 10.6±4.5 years [range, 3.0–17.9 years]) who were admitted to the hospital for acute abdomen or gastrointestinal bleeding and diagnosed with PUD by endoscopy and who underwent TUS were included. Clinical data were retrospectively collected by reviewing patient medical records. Gastric ulcer (GU) was suspected when the gastric wall exceeded 8 mm in thickness and had lost its five-layer structure on TUS. Duodenal ulcer (DU) was suspected if the duodenal wall thickness exceeded 5 mm. RESULTS: Sensitivity of TUS in diagnosing PUD was 66.7% for GU and 38.9% for DU. Mean age and body weight of the 11 patients suspected with PUD on TUS were 10.9±4.4 years and 38.1±17.2 kg, respectively. For 13 patients without suspected PUD, they were 12.1±4.1 years and 39.6±17.0 kg, respectively. There was a significant difference in height, weight, and body mass index between patients who were suspected to have PUD and those who were not suspected on TUS (p=0.014, 0.008, and 0.005, respectively). A significant difference in the sensitivity of TUS in diagnosing PUD was found between patients under 30 kg and those over 30 kg (88.9% and 20.0%, respectively; p=0.003). CONCLUSION: TUS investigation of the stomach and duodenum is an efficient method for PUD detection in children with low body weight. TUS can be used in preliminary diagnostic work-up before further invasive tests.
Abdomen, Acute ; Body Mass Index ; Body Weight ; Child* ; Diagnosis* ; Duodenal Ulcer ; Duodenum ; Endoscopy ; Female ; Hemorrhage ; Humans ; Medical Records ; Methods ; Peptic Ulcer* ; Retrospective Studies ; Stomach ; Stomach Ulcer ; Ultrasonography*

BACKGROUND/AIMS: Gastrointestinal (GI) bleeding is a life-threatening complication in critically ill patients. The aim of this study was to determine the efficacy of bedside endoscopy in an intensive care unit (ICU) setting, and to compare the outcomes of early endoscopy (within 24 hours of detecting GI bleeding) with late endoscopy (after 24 hours). METHODS: We retrospectively reviewed the medical records of patients who underwent bedside endoscopy for nonvariceal upper GI bleeding and lower GI bleeding that occurred after ICU admission at Seoul National University Hospital from January 2010 to May 2015. RESULTS: Two hundred and fifty-three patients underwent bedside esophagogastroduodenoscopy (EGD) for upper GI bleeding (early, 187; late, 66) and 69 underwent bedside colonoscopy (CS) for lower GI bleeding (early, 36; late, 33). Common endoscopic findings were peptic ulcer, and acute gastric mucosal lesion in the EGD group, as well as ischemic colitis and acute hemorrhagic rectal ulcers in the CS group. Early EGD significantly increased the rate of finding the bleeding focus (82% vs. 73%, p = 0.003) and endoscopic hemostasis (32% vs. 12%, p = 0.002) compared with late EGD. However, early CS significantly decreased the rate of identifying the bleeding focus (58% vs. 82%, p = 0.008) and hemostasis (19% vs. 49%, p = 0.011) compared with late CS due to its higher rate of poor bowel preparation and blood interference (38.9% vs. 6.1%, p = 0.035). CONCLUSIONS: Early EGD may be effective for diagnosis and hemostatic treatment in ICU patients with GI bleeding. However, early CS should be carefully performed after adequate bowel preparation.
Colitis, Ischemic ; Colonoscopy ; Critical Illness* ; Diagnosis ; Endoscopy* ; Endoscopy, Digestive System ; Hemorrhage* ; Hemostasis ; Hemostasis, Endoscopic ; Humans ; Intensive Care Units ; Medical Records ; Peptic Ulcer ; Retrospective Studies ; Seoul ; Ulcer

BACKGROUND/AIMS: Rebleeding is associated with mortality in patients with peptic ulcer bleeding (PUB), and risk stratification is important for the management of these patients. The purpose of our study was to examine the risk factors associated with rebleeding in patients with PUB. METHODS: The Korean Peptic Ulcer Bleeding registry is a large prospectively collected database of patients with PUB who were hospitalized between 2014 and 2015 at 28 medical centers in Korea. We examined the basic characteristics and clinical outcomes of patients in this registry. Univariate and multivariate analyses were performed to identify the factors associated with rebleeding. RESULTS: In total, 904 patients with PUB were registered, and 897 patients were analyzed. Rebleeding occurred in 7.1% of the patients (64), and the 30-day mortality was 1.0% (nine patients). According to the multivariate analysis, the risk factors for rebleeding were the presence of co-morbidities, use of multiple drugs, albumin levels, and hematemesis/hematochezia as initial presentations. CONCLUSIONS: The presence of co-morbidities, use of multiple drugs, albumin levels, and initial presentations with hematemesis/hematochezia can be indicators of rebleeding in patients with PUB. The wide use of proton pump inhibitors and prompt endoscopic interventions may explain the low incidence of rebleeding and low mortality rates in Korea.
Hemorrhage* ; Humans ; Incidence ; Korea ; Mortality ; Multivariate Analysis ; Peptic Ulcer Hemorrhage ; Peptic Ulcer* ; Prospective Studies ; Proton Pump Inhibitors ; Risk Factors