Background: s/Aims: Artificial intelligence (AI) technology has been used to assess surgery quality, educate, and evaluate surgical performance using video recordings in the minimally invasive surgery era. Much attention has been paid to automating surgical workflow analysis from surgical videos for an effective evaluation to achieve the assessment and evaluation. This study aimed to design a deep learning model to automatically identify surgical phases using laparoscopic cholecystectomy videos and automatically assess the accuracy of recognizing surgical phases. Methods: One hundred and twenty cholecystectomy videos from a public dataset (Cholec80) and 40 laparoscopic cholecystectomy videos recorded between July 2022 and December 2022 at a single institution were collected. These datasets were split into training and testing datasets for the AI model at a 2:1 ratio. Test scenarios were constructed according to structural characteristics of the trained model. No pre- or post-processing of input data or inference output was performed to accurately analyze the effect of the label on model training. Results: A total of 98,234 frames were extracted from 40 cases as test data. The overall accuracy of the model was 91.2%. The most accurate phase was Calot’s triangle dissection (F1 score: 0.9421), whereas the least accurate phase was clipping and cutting (F1 score:0.7761). Conclusions Our AI model identified phases of laparoscopic cholecystectomy with a high accuracy.

Background: s/Aims: Artificial intelligence (AI) technology has been used to assess surgery quality, educate, and evaluate surgical performance using video recordings in the minimally invasive surgery era. Much attention has been paid to automating surgical workflow analysis from surgical videos for an effective evaluation to achieve the assessment and evaluation. This study aimed to design a deep learning model to automatically identify surgical phases using laparoscopic cholecystectomy videos and automatically assess the accuracy of recognizing surgical phases. Methods: One hundred and twenty cholecystectomy videos from a public dataset (Cholec80) and 40 laparoscopic cholecystectomy videos recorded between July 2022 and December 2022 at a single institution were collected. These datasets were split into training and testing datasets for the AI model at a 2:1 ratio. Test scenarios were constructed according to structural characteristics of the trained model. No pre- or post-processing of input data or inference output was performed to accurately analyze the effect of the label on model training. Results: A total of 98,234 frames were extracted from 40 cases as test data. The overall accuracy of the model was 91.2%. The most accurate phase was Calot’s triangle dissection (F1 score: 0.9421), whereas the least accurate phase was clipping and cutting (F1 score:0.7761). Conclusions Our AI model identified phases of laparoscopic cholecystectomy with a high accuracy.

Background: s/Aims: Artificial intelligence (AI) technology has been used to assess surgery quality, educate, and evaluate surgical performance using video recordings in the minimally invasive surgery era. Much attention has been paid to automating surgical workflow analysis from surgical videos for an effective evaluation to achieve the assessment and evaluation. This study aimed to design a deep learning model to automatically identify surgical phases using laparoscopic cholecystectomy videos and automatically assess the accuracy of recognizing surgical phases. Methods: One hundred and twenty cholecystectomy videos from a public dataset (Cholec80) and 40 laparoscopic cholecystectomy videos recorded between July 2022 and December 2022 at a single institution were collected. These datasets were split into training and testing datasets for the AI model at a 2:1 ratio. Test scenarios were constructed according to structural characteristics of the trained model. No pre- or post-processing of input data or inference output was performed to accurately analyze the effect of the label on model training. Results: A total of 98,234 frames were extracted from 40 cases as test data. The overall accuracy of the model was 91.2%. The most accurate phase was Calot’s triangle dissection (F1 score: 0.9421), whereas the least accurate phase was clipping and cutting (F1 score:0.7761). Conclusions Our AI model identified phases of laparoscopic cholecystectomy with a high accuracy.

Background: The Korean government has established Health Plan and reinforced tobacco control policies step by step according to Framework Convention on Tobacco Control (FCTC). This study aims to investigate yearly smoking and secondhand smoking (SHS) exposure rates adjusted by demographic and socioeconomic factors. Methods: Multiple logistic regression analysis was conducted about the smoking experience, current smoking, and exposure to SHS at home, at work, and in public places using data from the 6th to 8th Korea National Health and Nutrition Examination Survey (2015– 2020). Results: Comparing 2015 with other years, smoking experience rates significantly decreased from 2015 to 2018 in men and significantly increased from 2018 to 2020 in women. Compared to 2015, current smoking rates significantly decreased only in 2020 for men and significantly increased in 2018 and 2019 for women. The rate of exposure to SHS at home significantly decreased until 2018. Rates of exposure to SHS in the workplace and public places were significant all year. Conclusion There were a slowly decreasing trend in men and an increasing trend in women at current smoking rates. It was found that there were significant decreasing trends at exposure to SHS. There is a need to set policies that reduce current smoking rates and especially to set a customized program to lower women’s smoking rates.

Background/Aims: Nonalcoholic fatty liver disease (NAFLD) is closely related to gut-microbiome. There is a paucity of research on which strains of gut microbiota affect the progression of NAFLD. This study explored the NAFLD-associated microbiome in humans and the role of Lactobacillus in the progression of NAFLD in mice. Methods: The gut microbiome was analyzed via next-generation sequencing in healthy people (n=37) and NAFLD patients with elevated liver enzymes (n=57). Six-week-old male C57BL/6J mice were separated into six groups (n=10 per group; normal, Western, and four Western diet + strains [109 colony-forming units/g for 8 weeks; L. acidophilus, L. fermentum, L. paracasei, and L. plantarum]). Liver/body weight ratio, liver pathology, serum analysis, and metagenomics in the mice were examined. Results: Compared to healthy subjects (1.6±4.3), NAFLD patients showed an elevated Firmicutes/Bacteroidetes ratio (25.0±29.0) and a reduced composition of Akkermansia and L. murinus (P<0.05). In the animal experiment, L. acidophilus group was associated with a significant reduction in liver/body weight ratio (5.5±0.4) compared to the Western group (6.2±0.6) (P<0.05). L. acidophilus (41.0±8.6), L. fermentum (44.3±12.6), and L. plantarum (39.0±7.6) groups showed decreased cholesterol levels compared to the Western group (85.7±8.6) (P<0.05). In comparison of steatosis, L. acidophilus (1.9±0.6), L. plantarum (2.4±0.7), and L. paracasei (2.0±0.9) groups showed significant improvement of steatosis compared to the Western group (2.6±0.5) (P<0.05). Conclusions Ingestion of Lactobacillus, such as L. acidophilus, L. fermentum, and L. plantarum, ameliorates the progression of nonalcoholic steatosis by lowering cholesterol. The use of Lactobacillus can be considered as a useful strategy for the treatment of NAFLD.

Background/Aims: Vascular disease is an established risk factor for stroke in patients with atrial fibrillation (AF), which is included in CHA2DS2-VASc score. However, the role of carotid atherosclerosis remains to be determined. Methods: Three hundred-ten patients with AF who underwent carotid sonography were enrolled. Results: During a median follow-up of 31 months, 18 events (5.8%) of stroke were identified. Patients with stroke had higher carotid intima-media thickness (CIMT) (1.16 ± 0.33 mm vs. 0.98 ± 0.25 mm, p = 0.017). CIMT was significantly increased according to the CHA2DS2-VASc score (p < 0.001) and it was correlated with left ventricular mass index and early diastolic mitral annular velocity (e’), a ratio of early transmitral flow velocity to e’ (E/e’) and pulmonary artery systolic pressure (all p < 0.05). Cox regression using multivariate models showed that carotid plaque was associated with the risk of stroke (hazard ratio, 3.748; 95% confidence interval [CI], 1.107 to 12.688; p = 0.034). C-statistics increased from 0.648 (95% CI, 0.538 to 0.757) to 0.716 (95% CI, 0.628 to 0.804) in the CHA2DS2-VASc score model after the addition of CIMT and carotid plaque as a vascular component (p = 0.013). Conclusions Increased CIMT and presence of carotid plaque are associated with a high risk of ischemic stroke, and CIMT is related to myocardial remodeling and diastolic dysfunction, suggesting that carotid atherosclerosis can improve risk prediction of stroke in patients with AF, when included under vascular disease in the CHA2DS2-VASc scoring system.

The aim of this study was to assess the potential of a prone chest CT for the evaluation of esophageal cancer, as compared with a routine supine chest CT. 69 patients (67±18 years old) with pathologically confirmed esophageal cancers underwent MDCT in the supine and prone positions. The supine CT was performed first, followed by the prone position. Localization and staging of individual esophageal lesions on both the prone and supine CTs were assessed by two thoracic radiologists, using a scoring system that consisted of three confidence scales, and the results were correlated with the endoscopic and surgical findings. The mean confidence score for the detection of esophageal cancer was higher in the prone position (2.58±0.74) than that in the supine position (2.42±0.83) with statistical significance (p=0.002). The mean confidence score for predicting local invasion in the selected patients (n=18) who underwent esophagectomy was also higher in the prone position (2.39±0.85) than that in the supine position (2.06±0.73) with statistical significance (p=0.01). In 10 of 11 cases that showed definitive determination for periesophageal infiltration or adjacent organ invasion on the prone CT (score 3), the corresponding findings were also observed in the post-operative evaluations. In conclusion, prone chest CT for evaluating esophageal cancer could have advantages in regards to the localization of esophageal cancer and predicting local invasion compared to that of routine supine CT and can improve the diagnostic accuracy of chest CTs.

In the published article, the Figure 4 was published with incorrect y-axis and legend.

Procalcitonin (PCT) is a predictive marker for the occurrence of bacterial infection and the decision to terminate antibiotic treatment in critically ill patients. An unusual increase in PCT, regardless of infection, has been observed during extracorporeal membrane oxygenation (ECMO) support. We evaluated trends and the predictive value of PCT levels in adult cardiogenic shock during treatment with ECMO. We reviewed the clinical records of 38 adult cardiogenic shock patients undergoing veno-arterial ECMO support between January 2014 and December 2016. The exclusion criteria were age < 18 years, pre-ECMO infection, and less than 48 hours of support. The mean patient age was 56.7±14.7 years and 12 (31.6%) patients were female. The mean duration of ECMO support was 9.0±7.6 days. The rates of successful ECMO weaning and survival to discharge were 55.3% (n=21) and 52.6% (n=20), respectively. There were 17 nosocomial infections in 16 (42.1%) patients. Peak PCT levels (mean 25.6±9.4 ng/mL) were reached within 48 hours after initiation of ECMO support and decreased to ≤5 ng/mL within one week. The change in PCT levels was not useful in predicting the occurrence of new nosocomial infections during the ECMO run. However, a PCT level >10 ng/mL during the first week of ECMO support was significantly associated with mortality (p < 0.01). The change in PCT level was not useful in predicting new infection during ECMO support. However, higher PCT levels within the first week of the ECMO run are associated with significantly higher mortality.
Adult ; Bacterial Infections ; Calcitonin ; Critical Illness ; Cross Infection ; Extracorporeal Membrane Oxygenation ; Female ; Humans ; Mortality ; Shock ; Shock, Cardiogenic ; Weaning

BACKGROUND: It has been reported that heat shock protein 70 (HSP70) and interleukin-8 (IL-8) play an important role in cells during the wound healing process. However, there has been no report on the effect of HSP70 and IL-8 on the blisters of burn patients. OBJECTIVE: This study aimed to evaluate the serial quantitative changes of HSP70 and IL-8 in burn blisters. METHODS: Twenty-five burn patients were included, for a total of 36 cases: twenty cases on the first day, six cases on the second, five cases on the third, three cases on the fourth, and two cases on the fifth. A correlation analysis was performed to determine the relationship between the concentration of HSP70 and IL-8 and the length of the treatment period. RESULTS: The HSP70 concentration was the highest on the first day, after which it decreased down to near zero. Most HSP70 was generated during the first 12 hours after the burn accident. There was no correlation between the concentration of HSP70 on the first day and the length of the treatment period. No measurable concentration of IL-8 was detected before 5 hours, but the concentration started to increase after 11 hours. The peak value was measured on the fourth day. CONCLUSION: While HSP70 increased in the first few hours and decreased afterwards, IL-8 was produced after 11 hours and increased afterward in burn blister fluid. These findings provide new evidence on serial changes of inflammatory mediators in burn blister fluid.
Blister* ; Burns* ; Heat-Shock Proteins* ; Hot Temperature* ; HSP70 Heat-Shock Proteins* ; Humans ; Interleukin-8 ; Wound Healing ; Wounds and Injuries