Development and Validation of Deep-Learning Algorithm for Electrocardiography-Based Heart Failure Identification

BACKGROUND AND OBJECTIVES@#Screening and early diagnosis for heart failure (HF) are critical. However, conventional screening diagnostic methods have limitations, and electrocardiography (ECG)-based HF identification may be helpful. This study aimed to develop and validate a deep-learning algorithm for ECG-based HF identification (DEHF).@*METHODS@#The study involved 2 hospitals and 55,163 ECGs of 22,765 patients who performed echocardiography within 4 weeks were study subjects. ECGs were divided into derivation and validation data. Demographic and ECG features were used as predictive variables. The primary endpoint was detection of HF with reduced ejection fraction (HFrEF; ejection fraction [EF]≤40%), and the secondary endpoint was HF with mid-range to reduced EF (≤50%). We developed the DEHF using derivation data and the algorithm representing the risk of HF between 0 and 1. We confirmed accuracy and compared logistic regression (LR) and random forest (RF) analyses using validation data.@*RESULTS@#The area under the receiver operating characteristic curves (AUROCs) of DEHF for identification of HFrEF were 0.843 (95% confidence interval, 0.840–0.845) and 0.889 (0.887–0.891) for internal and external validation, respectively, and these results significantly outperformed those of LR (0.800 [0.797–0.803], 0.847 [0.844–0.850]) and RF (0.807 [0.804–0.810], 0.853 [0.850–0.855]) analyses. The AUROCs of deep learning for identification of the secondary endpoint was 0.821 (0.819–0.823) and 0.850 (0.848–0.852) for internal and external validation, respectively, and these results significantly outperformed those of LR and RF.@*CONCLUSIONS@#The deep-learning algorithm accurately identified HF using ECG features and outperformed other machine-learning methods.

Development and Validation of Deep-Learning Algorithm for Electrocardiography-Based Heart Failure Identification

BACKGROUND AND OBJECTIVES: Screening and early diagnosis for heart failure (HF) are critical. However, conventional screening diagnostic methods have limitations, and electrocardiography (ECG)-based HF identification may be helpful. This study aimed to develop and validate a deep-learning algorithm for ECG-based HF identification (DEHF). METHODS: The study involved 2 hospitals and 55,163 ECGs of 22,765 patients who performed echocardiography within 4 weeks were study subjects. ECGs were divided into derivation and validation data. Demographic and ECG features were used as predictive variables. The primary endpoint was detection of HF with reduced ejection fraction (HFrEF; ejection fraction [EF]≤40%), and the secondary endpoint was HF with mid-range to reduced EF (≤50%). We developed the DEHF using derivation data and the algorithm representing the risk of HF between 0 and 1. We confirmed accuracy and compared logistic regression (LR) and random forest (RF) analyses using validation data. RESULTS: The area under the receiver operating characteristic curves (AUROCs) of DEHF for identification of HFrEF were 0.843 (95% confidence interval, 0.840–0.845) and 0.889 (0.887–0.891) for internal and external validation, respectively, and these results significantly outperformed those of LR (0.800 [0.797–0.803], 0.847 [0.844–0.850]) and RF (0.807 [0.804–0.810], 0.853 [0.850–0.855]) analyses. The AUROCs of deep learning for identification of the secondary endpoint was 0.821 (0.819–0.823) and 0.850 (0.848–0.852) for internal and external validation, respectively, and these results significantly outperformed those of LR and RF. CONCLUSIONS: The deep-learning algorithm accurately identified HF using ECG features and outperformed other machine-learning methods.

Smart Device Usage-Related Factors are Correlated with Self-Regulation Ability in Early Childhood / 대한소아신경학회지

PURPOSE: To investigate the relationship between the smart devices usage-related factors and self-regulation ability development in early childhood. METHODS: Parental questionnaires of 187 children aged 3–6 years were analyzed. The metrics included smart device usage frequency (times/week, scored as uFreq), smart device usage time (hours/day, scored as uTime), parental scale for appropriate smart device usage level (scored as uLevel), the Korean-developmental screening test (K-DST), and the scale for self-regulation ability in young children (scored as SRS, and including four sub-categories: self-appraisal, self-determination, behavior inhibition, and emotionality). The correlations were analyzed by total age group and by each age. RESULTS: In the total age group analysis, uFreq and uTime were negatively correlated with mean SRS (rs =−0.366, −0.330; P < 0.001) and sub-category SRS (rs =−0.186 to −0.370; P < 0.05). Mean uLevel score was positively correlated with mean SRS (rs =0.406; P < 0.001) and most of the mean sub-category SRS (rs =0.174 to 0.362; P < 0.05). In 3-year-old children, the mean SRS was strongly negatively correlated with uFreq (rs =−0.751; P < 0.001), negatively correlated with uTime (rs =−0.518; P < 0.001), and positively correlated with mean uLevel score (rs =0.533; P=0.013). Such correlations seemed to decrease at the age of 4–6 years. CONCLUSION: Self-regulation ability was significantly correlated with smart device-related factors and was the highest in the 3-year-old children. Encouraging appropriate smart device usage will be helpful for self-regulation development of young children.

Effectiveness of bee venom acupuncture in alleviating post-stroke shoulder pain: a systematic review and meta-analysis / 中西医结合学报

<p><b>BACKGROUND</b>Shoulder pain is a common complication of stroke. Bee venom acupuncture (BVA) is increasingly used in the treatment of post-stroke shoulder pain.</p><p><b>OBJECTIVE</b>To summarize and evaluate evidence on the effectiveness of BVA in relieving shoulder pain after stroke.</p><p><b>SEARCH STRATEGY</b>Nine databases, namely MEDLINE, EMBASE, the Cochrane Library, the China National Knowledge Infrastructure (CNKI), the Japan Science and Technology Information Aggregator, Electronic (J-STAGE), and four Korean medical databases, namely, the National Assembly Library, the Research Information Service System, the National Discovery for Science Leaders, and OASIS, were searched from their inception through August 2014 without language restrictions.</p><p><b>INCLUSION CRITERIA</b>Randomized controlled trials (RCTs) were included if BVA was used at acupoints as the sole treatment, or as an adjunct to other treatments, for shoulder pain after stroke.</p><p><b>DATA EXTRACTION AND ANALYSIS</b>Two review authors independently selected trials for inclusion, assessed methodological quality and extracted data.</p><p><b>RESULTS</b>A total of 138 potentially relevant articles were identified, 4 of which were RCTs that met our inclusion criteria. The quality of studies included was generally low, and a preponderance of positive results was demonstrated. All four trials reported favorable effects of BVA on shoulder pain after stroke. Two RCTs assessing the effects of BVA on post-stroke shoulder pain, as opposed to saline injections, were included in the meta-analysis. Pain was significantly lower for BVA than for saline injections (standardized mean difference on 10-cm visual analog scale: 1.46 cm, 95% CI=0.30-2.62, P=0.02, n=86) CONCLUSION: This review provided evidence suggesting that BVA is effective in relieving shoulder pain after stroke. However, further studies are needed to confirm the role of BVA in alleviating post-stroke shoulder pain. Future studies should be conducted with large samples and rigorous study designs.</p>

Gastric duplication cyst associated with dorsal pancreatic agenesis and a cystic pulmonary lesion / 대한내과학회지

Gastric duplication cysts and dorsal pancreatic agenesis are rare congenital anomalies, and little is known of the association between these two anomalies. A 17-year-old woman was admitted with a cystic lung mass detected as part of a health screening program. Chest computed tomography (CT) showed a cystic lung lesion in the right lower lobe and an incidental cystic abdominal mass. She had no specific symptoms or signs. Abdominal CT and endoscopic ultrasonography (EUS) revealed a 9x7x5.5-cm cystic mass and agenesis of the body and tail of the pancreas. The cyst was removed successfully using laparoscopic gastric wedge resection. Microscopically, the muscle coat of the cyst wall was fused with the muscle layer of the stomach. Therefore, the cystic mass was diagnosed as a gastric duplication cyst. We present a patient with a pulmonary cystic lesion associated with two congenital anomalies: a gastric duplication cyst and dorsal pancreatic agenesis.

Hepatitis B virus reactivation during chlorambucil and prednisolone treatment in an HBsAg-negative and anti-HBs-positive patient with B-cell chronic lymphocytic leukemia / 대한간학회지

It is generally accepted that seroconversion of hepatitis B virus (HBV) surface antigen (HBsAg) to an antibody to HBsAg (anti-HBs) indicates clearance of HBV. Here we report a case of severe hepatitis that manifested during chemotherapy in a female patient with chronic lymphocytic leukemia (CLL) who had been initially seronegative for HBsAg and seropositive for anti-HBs. The patient received chlorambucil and prednisolone for the treatment of CLL. After 6 months the serum levels of aminotransferases were increased, and HBsAg and HBV DNA were present in serum. Lamivudine was administered immediately after confirming the HBV reactivation, which considerably improved jaundice and aminotransferase levels after 3 weeks. The patient was able to resume the chemotherapy whilst continuing lamivudine treatment. This case report highlights the need for physicians to be aware of the potential risk of HBV reactivation even in an HBsAg-negative person but with detectable anti-HBc and/or anti-HBs, underscoring the need for future studies that explore the role of antiviral prophylaxis in this setting.

Clinical Implication of Natriuretic Peptides and Left Atrial Volume for the Screening Methods of Advanced Diastolic Dysfunction in the Community

BACKGROUND: Recently, B-natriuretic peptide (BNP) level and left atrial volume index (LAVi) were known to correlate with indices of LV diastolic function. As a screening method, we tried to evaluate the efficacy to BNP, ANP, and LAVi to predict the advanced diastolic dysfunction that means myocardial relaxation abnormality and elevated LV filling pressure. METHODS: In 100 patients who referred for echocardiography, Doppler recording of the mitral inflow and tissue Doppler imaging of the mitral annulus were obtained and classified into 4 diastolic function grades (normal, impaired relaxation, pseudonormal, and restrictive). Advanced diastolic dysfunction was defined as pseudonormal and restrictive physiology. LAVi was measured by modified Simpson's method in apical 4-chamber view at end-systole. Plasma levels of BNP and ANP were measured on the same day as echocardiogram was done. RESULTS: BNP and ANP levels were increased as diastolic function grade was worsening (BNP : 60+/-92, 108+/-204, 778+/-1,023 and 1,426+/-1,421 pg/ml, p<0.001; ANP: 22+/-30, 23+/-26, 94+/-92, 96+/-61 pg/ml, p<0.001). LAVi was also increased as diastolic dysfunction was advanced: 24+/-7 ml/m2, 27+/-9 ml/m2, 37+/-12 ml/m2, 45+/-12 ml/m2, p<0.001. The areas under the curve of receiver-operator characteristic curve for BNP, ANP and LAVi to detect the advanced diastolic dysfunction were 0.91, 0.88 and 0.84, respectively. BNP of 137 pg/ml, ANP of 34 pg/ml, and LAVi of 30 ml/m2 were the best values of sensitivity and specificity, respectively. CONCLUSION: These data suggest that BNP, ANP and LAVi provide meaningful sensitivity and specificity for the detection of advanced diastolic dysfunction, respectively. Among these, BNP is better than ANP or LAVi for the screening method to predict the advanced diastolic dysfunction.

A Case of Ethylene Glycol Poisoning with Metabolic Acidosis Treated with Hemodialysis / 대한신장학회잡지

We present a case of ethylene glycol poisoning with high anion gap metabolic acidosis. A 71 year-old female patient was transferred to our hospital after ingesting 450 mL of anti-freeze. At arrival she showed high anion gap metabolic acidosis with pH 7.035, PaCO2 7.2 mmHg, PaO2 117.5 mmHg, HCO3 - 1.9 mmol/L and anion gap 32 mmol/L. Calcium oxalate crystals were identified on urine microscopy. Bicarbonate treatment did not improve her metabolic acidosis, and oliguric acute renal failure was developed. So she was treated with hemodialysis. After the hemodialysis treatment her metabolic acidosis was corrected and her renal function was improved. She was discharged on the 22nd day.

A Case of Successful Endoscopic Therapy in Iatrogenic Perforation of the Colon during Colonoscopy / 대한소화기내시경학회지

Colonoscopy is a safe and standard procedure for diagnosis and therapy of colonic disorders. Iatrogenic colonic perforation during diagnostic colonoscopy, a rare abdominal emergency, has an 0.3~0.8% incidence rate. The choice of treatment for this complication remains controversial. Prompt operative intervention is preferred to minimize morbidity and mortality. However, operative intervention is invasive and needs a long-term recovery period. Conservative treatment is less invasive but can lead to more extensive surgery in case of treatment failure. Very important point on the treatment of iatrogenic perforation of the colon during diagnostic colonoscopy is to avoid the leaking of intestinal contents into the intraperitoneal cavity. We report here a case in which an iatrogenic perforation of the colon during diagnostic colonoscopy was successfully treated by endoscopic clip therapy.