Improving difficult direct laryngoscopy prediction using deep learning and minimal image analysis: a single-center prospective study.

Accurate prediction of difficult direct laryngoscopy (DDL) is essential to ensure optimal airway management and patient safety. The present study proposed an AI model that would accurately predict DDL using a small number of bedside pictures of the patient's face and neck taken simply with a smartphone. In this prospective single-center study, adult patients scheduled for endotracheal intubation under general anesthesia were included. Patient pictures were obtained in frontal, lateral, frontal-neck extension, and open mouth views. DDL prediction was performed using a deep learning model based on the EfficientNet-B5 architecture, incorporating picture view information through multitask learning. We collected 18,163 pictures from 3053 patients. After under-sampling to achieve a 1:1 image ratio of DDL to non-DDL, the model was trained and validated with a dataset of 6616 pictures from 1283 patients. The deep learning model achieved a receiver operating characteristic area under the curve of 0.81-0.88 and an F1-score of 0.72-0.81 for DDL prediction. Including picture view information improved the model's performance. Gradient-weighted class activation mapping revealed that neck and chin characteristics in frontal and lateral views are important factors in DDL prediction. The deep learning model we developed effectively predicts DDL and requires only a small set of patient pictures taken with a smartphone. The method is practical and easy to implement.

Association between COVID-19 and incidence of cardiovascular disease and all-cause mortality among patients with diabetes.

Introduction: Although the risk of coronavirus disease 2019 (COVID-19) infection is higher in patients who are diagnosed with diabetes than in those who are not, research on the risk of cardiovascular disease (CVD) in COVID-19 infected patients diagnosed with diabetes compared to those who are not infected by COVID-19 is lacking. This study aimed to examine the association between COVID-19, incidence of CVD, and all-cause mortality in patients with diabetes. Methods: This study used data from the Health Insurance Review and Assessment, and included 16,779 patients with COVID-19 and 16,779 matched controls between January 2017 and June 2021. The outcomes included cardiovascular disease (CVD), coronary heart disease, stroke, and all-cause mortality. Cox proportional hazards regression models were used to evaluate these associations. Results: Patients with diabetes hospitalized because of COVID-19 had a significantly increased risk of CVD (adjusted hazard ratio [AHR], 2.12; 95% confidence interval [CI]: 1.97, 2.27) than those without COVID-19. The risks of coronary heart disease (AHR, 2.00; 95% CI: 1.85, 2.17) and stroke (AHR, 2.21; 95% CI: 1.90, 2.57) were higher in the intervention group than in the control group. In the case of all-cause mortality for middle-aged adults, we observed a higher risk in diabetes patients hospitalized due to COVID-19 than in patients without COVID-19 (AHR, 1.37; 95% CI: 1.18, 1.59). Conclusions: This study showed that patients with diabetes hospitalized due to COVID-19 had an increased risk of CVD, coronary heart disease, stroke incidence, and mortality than those who were not COVID-19 infected, suggesting more careful prevention and management among patients with COVID-19.

Development and Validation of Deep Learning-Based Algorithms for Predicting Lumbar Herniated Nucleus Pulposus Using Lumbar X-rays.

Lumbar herniated nucleus pulposus (HNP) is difficult to diagnose using lumbar radiography. HNP is typically diagnosed using magnetic resonance imaging (MRI). This study developed and validated an artificial intelligence model that predicts lumbar HNP using lumbar radiography. A total of 180,271 lumbar radiographs were obtained from 34,661 patients in the form of lumbar X-ray and MRI images, which were matched together and labeled accordingly. The data were divided into a training set (31,149 patients and 162,257 images) and a test set (3512 patients and 18,014 images). Training data were used for learning using the EfficientNet-B5 model and four-fold cross-validation. The area under the curve (AUC) of the receiver operating characteristic (ROC) for the prediction of lumbar HNP was 0.73. The AUC of the ROC for predicting lumbar HNP in L (lumbar) 1-2, L2-3, L3-4, L4-5, and L5-S (sacrum)1 levels were 0.68, 0.68, 0.63, 0.67, and 0.72, respectively. Finally, an HNP prediction model was developed, although it requires further improvements.

Obesity increases the risk of erosive esophagitis but metabolic unhealthiness alone does not: a large-scale cross-sectional study.

BACKGROUND: Obesity is a known risk factor for erosive esophagitis (EE) and metabolic unhealthiness has been implicated in EE pathogenesis. However, obesity and metabolic unhealthiness are not synonymous and the associations between obesity, metabolic health, and EE are unclear. Therefore, our aim was to investigate the relationship between EE, obesity, and metabolic health. METHODS: We performed a retrospective cross-sectional study of subjects undergoing health screening at a university hospital. Subjects were classified into 4 groups based on metabolic and obesity criteria: metabolically healthy nonobese (MHNO), metabolically healthy obese (MHO), metabolically unhealthy nonobese (MUNO), and metabolically unhealthy obese (MUO). Multivariable analysis was used to identify EE risk factors with MHNO subjects as reference. To determine if there were synergistic interactions between metabolic health and obesity status, the Rothman's synergy index and attributable proportion of risk were also calculated. RESULTS: We included 10,338 subjects (5448 MHNO, 1605 MHO, 1600 MUNO, 1685 MUO). The prevalence of EE was 6.5% in MHNO, 12.6% in MHO, 9.3% in MUNO, and 14.3% in MUO. EE risk was increased significantly by obesity (MHO: OR, 1.589, 95% CI, 1.314-1.921, P < 0.001; MUO: OR, 1.734, 95% CI, 1.441-2.085, P < 0.001), but not in MUNO subjects (OR, 1.224, 95% CI, 0.991-1.511, P = 0.060). Male sex, blood leukocyte count, alcohol, and smoking significantly increased EE risk, but H. pylori infection was protective. Replacement of obesity with abdominal obesity gave similar results. The Rothman's synergy index was 0.920 (95% CI, 0.143-5.899) and the attributable proportion of risk was - 0.051 (95% CI, - 1.206-1.105), indicating no interaction between metabolic and obesity status on EE risk. CONCLUSIONS: We demonstrated that obesity increased the risk of EE, regardless of metabolic health status. However, EE risk was not significantly increased in MUNO subjects, suggesting that metabolic unhealthiness may not be involved in EE pathogenesis. As observational cross-sectional studies cannot prove causality, prospective longitudinal studies involving obesity and metabolic treatment should be performed to further investigate the association between obesity, metabolic health, and EE risk.

Obesity and Metabolic Unhealthiness Have Different Effects on Colorectal Neoplasms.

Context: Obesity and insulin resistance are risk factors for colorectal neoplasms (CRN), but data regarding metabolic status, obesity, and CRN are lacking. Objective: To investigate the relationship between metabolic status, obesity, and CRN in Koreans who underwent colonoscopy. Design: Retrospective, cross-sectional. Participants: Subjects were divided based on metabolic and obesity criteria, as follows: metabolically healthy nonobese (MHNO), metabolically healthy obese (MHO), metabolically unhealthy nonobese (MUNO), and metabolically unhealthy obese (MUO). Main Outcome Measures: Multiple regression was used to identify CRN and advanced CRN risk factors, with the MHNO group as reference. Results: A total of 10,235 subjects was included, as follows: 5096 MHNO, 1538 MHO, 1746 MUNO, and 1855 MUO. Of these, 3297 had CRN (32.2%), and 434 (4.2%) had advanced CRN. Number of subjects with CRN in each group were: MHNO 25.8%, MHO 33.9%, MUNO 38.9%, and MUO 42.0% (P for trend < 0.001). Risk of CRN was increased in the MHO [odds ratio (OR) 1.239, 95% confidence interval (CI) 1.082 to 1.418, P = 0.002], the MUNO (OR 1.233, 95% CI 1.086 to 1.400, P = 0.001), and the MUO groups (OR 1.510, 95% CI 1.338 to 1.706, P < 0.001), whereas risk of advanced CRN was increased in the MUNO (OR 1.587, 95% CI 1.222 to 2.062, P = 0.001) and the MUO groups (OR 1.456, 95% CI 1.116 to 1.900, P = 0.006). Conclusions: Obesity increased CRN risk with metabolically unhealthy status adding risk. For advanced CRN, metabolically unhealthy status increased the risk but obesity did not.

Clinical characteristics, risk factors and outcomes of asymptomatic rectal neuroendocrine tumors.

BACKGROUND: Rectal neuroendocrine tumors (NET) are often asymptomatic and frequently discovered during health examinations. However, data on the risk factors of asymptomatic rectal NETs are lacking. We investigated the risk factors, clinical characteristics and outcomes of asymptomatic rectal NETs discovered during health screening. MATERIALS AND METHODS: Asymptomatic subjects who underwent colonoscopy during routine health screening at a tertiary hospital from March 2009 to July 2014 were reviewed. Subjects with histologically confirmed rectal NETs were compared with healthy controls from the same population. Risk factors for rectal NETs were analyzed by multivariable analysis. Clinical outcomes of the resected NETs were also analyzed. RESULTS: A total of 21,706 Subjects underwent screening colonoscopy during the study period. 3417 were excluded from the study, and 180 rectal NET subjects were compared with 18,109 controls. Multivariable analysis showed that a previous history of malignancy (OR 2.960, 95% CI 1.673-5.237, p < 0.001), hypertriglyceridemia (OR 1.482, 95% CI 1.046-2.100, p = 0.027), higher fasting plasma glucose levels (OR 1.008, 95% CI 1.003-1.014, p = 0.001) and higher carcinoembryonic antigen levels (OR 1.019, 95% CI 1.003-1.035, p = 0.021) were significant risk factors while older age (OR 0.964, 95% CI 0.951-0.977, p < 0.001) was a preventive factor. One hundred and sixty nine subjects had endoscopic resection, five were treated by local surgery and six by radical surgery. Complete resection was achieved in 152 subjects. There were three cases of positive lymph nodes and one metastasis. Histology revealed four lymphatic, five vascular and two cases of perineural invasion. One hundred and fifty seven subjects were followed up for at least 1 year (median 38.6 months, 12-84 months). There were no recurrences during the follow-up period. CONCLUSIONS: Younger age, previous history of malignancy, higher fasting plasma glucose levels and hypertriglyceridemia are significantly associated with rectal NET risk.

Obesity is associated with increasing esophageal Acid exposure in korean patients with gastroesophageal reflux disease symptoms.

BACKGROUND/AIMS: Obesity is regarded as an important contributor to the increasing occurrence of gastroesophageal reflux disease. The aims of this study were to determine whether obesity is associated with gastroesophageal reflux in patients with gastroesophageal reflux disease and to identify the factors affecting increased acid exposure in obese patients. METHODS: We retrospectively analyzed the data of patients who underwent ambulatory 24-hour pH monitoring and esophageal manometry at Seoul St. Mary's Hospital. Obesity was classified according to the Asia-Pacific criteria. RESULTS: A total of 366 patients were analyzed; 18 were underweight, 152 normal weight, 104 overweight, and 92 obese. Obesity was more frequent in men and younger patients. The percentage time of pH < 4 in the total, upright, and postprandial periods was significantly higher in obese patients than in normal or underweight patients. The DeMeester score was also higher in obese patients. Body mass index correlated positively with reflux parameters. Multivariate analysis showed that being male and obesity were significantly associated with abnormal acid exposure (P < 0.005). The total lower esophageal sphincter length shortened as body mass index increased (P < 0.005). The gastroesophageal pressure gradient increased as body mass index increased (P < 0.05). CONCLUSIONS: Obesity is associated with increasing esophageal acid exposure. The mechanism responsible for the relationship between gastroesophageal reflux disease and obesity may be associated with shortening of the lower esophageal sphincter length and increasing the gastroesophageal pressure gradient.

[Usefulness of ¹8F-fluoro-2-deoxyglucose positron emission tomography in evaluation of gastric cancer stage].

BACKGROUND/AIMS: The usefulness of ¹8F-fluoro-2-deoxyglucose (FDG)-PET in detecting primary cancer, lymph node metastasis, and distant metastasis were studied in the gastric cancer patients. METHODS: The subjects were 392 gastric cancer patients who received FDG-PET and an abdominal CT test prior to surgery. The results of FDG-PET and CT were compared with the surgical and pathologic results. RESULTS: The primary site detection rate of FDG-PET was 74.4%, 50.3% in early gastric cancer and 92.0% in advanced gastric cancer. Detection rate was higher when tumors were larger than 3.5 cm, had deeper depth of invasion, and at a later stage (p<0.05, respectively). In multivariate analysis, tumor size, spread of tumor cells beyond the muscle layer (≥T2), and lymph node metastasis were statistically significant factors in primary site detection rate. The sensitivity, specificity, and positive predictive value of FDG-PET to lymph node metastasis were 59.6%, 88.8%, and 81.1% respectively, sensitivity being lower compared to CT while specificity and positive predictive value were higher. Sensitivity, specificity, and positive predictive value to distant metastasis were, respectively, 66.7%, 99.2%, and 88.0%, similar to CT. In 21 of the 392 patients (5.4%), synchronous double primary cancers were detected. CONCLUSIONS: In gastric cancer, usefullness of FDG-PET is limited to the advanced stage. Diagnostic value of this test was not superior to CT. However, FDG-PET may be useful in detecting synchronous double primary cancers.

Synthesis and characterization of platinum(II) complexes with 2,2'-bipyridine derivative supporting ligands.

The reaction of the [Pt(bpy-R)Cl2](bpy-R: R=H (2,2'-bipyridine); R=CH3 (4,4'-dimethyl-2,2'-biypridine (DM-bpy), 3,3'-5,5'-tertamethyl-2,2'-bipyridiyl (TM-bpy)) with 1,4-Bis(5'-2',2"-bipyridine)benzene (bpy-Ph-bpy) affords the following mono- and di-platinum complexes of [(bpy)Pt(bpy-Ph-bpy)][PF6]2 (1), [(bpy)Pt(bpy-Ph-bpy)Pt(bpy)])[PF6]4 (2), [(DM-bpy)Pt(bpy-Ph-bpy)])][PF6]2 (3), and [(TM-bpy)Pt(bpy-ph-bpy)[PF6]2 (4), respectively. These complexes were characterized by NMR, IR, UV/VIS, PL and cyclic voltammetry. The internal quantum yields of these platinum(II) complexes are very high (0.83-0.99) and these complexes emit light at deep blue regions (373-417 nm). The redox behavior of complexes 1 and 2 shows quasi-reversible process.

Discrimination of metastatic from acute osteoporotic compression spinal fractures with MR imaging.

A study was performed to determine which magnetic resonance (MR) imaging findings are useful in discrimination between metastatic compression fractures and acute osteoporotic compression fractures of the spine. The MR imaging findings in 27 patients with metastatic compression fractures and 55 patients with acute osteoporotic compression fractures were compared by using the chi(2) test. MR imaging findings suggestive of metastatic compression fractures were as follows: a convex posterior border of the vertebral body, abnormal signal intensity of the pedicle or posterior element, an epidural mass, an encasing epidural mass, a focal paraspinal mass, and other spinal metastases. MR imaging findings suggestive of acute osteoporotic compression fractures were as follows: a low-signal-intensity band on T1- and T2-weighted images, spared normal bone marrow signal intensity of the vertebral body, retropulsion of a posterior bone fragment, and multiple compression fractures. The signal intensity on fast spin-echo T2-weighted images obtained without fat suppression played little role in distinguishing between metastatic compression fractures and acute osteoporotic compression fractures.