Purpose: Early identification of patients at risk for acute respiratory failure (ARF) could help clinicians devise preventive strategies. Analyzing biosignals with artificial intelligence (AI) can uncover hidden information and variability within time series. We aimed to develop and validate AI models to predict ARF within 72 h after emergency department admission, primarily using highresolution biosignals collected within 4 h of arrival. Materials and Methods: Our AI model, built on convolutional recurrent neural networks, combines biosignal feature extraction and sequence modeling. The model was developed and internally validated with data from 5284 admissions [1085 (20.5%) positive for ARF], and externally validated using data from 144 admissions [7 (4.9%) positive for ARF] from another institution. We defined ARF as the application of advanced respiratory support devices. Results: Our AI model performed well in predicting ARF, achieving area under the receiver operating characteristic curve (AUROC) of 0.840 and 0.743 in internal and external validations, respectively. It outperformed the Modified Early Warning Score (MEWS) and XGBoost models built only with clinical variables. High predictive ability for mortality was observed, with AUROC up to 0.809. A 10% increase in AI prediction scores was associated with 1.44-fold and 1.42-fold increases in ARF risk and mortality risk, respectively, even after adjusting for MEWS and demographic variables. Conclusion Our AI model demonstrates high predictive accuracy and significant associations with clinical outcomes. Our AI model has the potential to promptly aid in triage decisions. Our study shows that using AI to analyze biosignals advances disease detection and prediction.

Purpose: To evaluate the relationship between distal fusion level in correction and fusion surgery for adolescent idiopathic scoliosis (AIS) and radiologic changes in the sacroiliac (SI) joint. Materials and Methods: This retrospective cohort study evaluated patients who underwent correction and fusion for AIS between 2005 and 2017 with at least 5 years of follow-up. We categorized patients into two groups: Group 1 (distal fusion above L2, 74 patients) and Group 2 (distal fusion at L3 and below, 52 patients). Radiologic parameters and SI joint changes were evaluated on plain radiographs obtained from preoperative to 5 years postoperatively. We also investigated other risk factors for SI joint change. Results: Analysis of demographic factors revealed no significant difference between the two groups. There was a significant difference in the incidence of SI joint change between Group 1 (5 patients, 6.75%) and Group 2 (18 patients, 34.61%), with Group 2 showing a faster increase in incidence according to the Kaplan-Meier method (p<0.0001). Preoperative lumbar lordosis (LL) and ΔLL had a significant relationship with SI joint changes [preoperative LL, hazard ratio (HR)=0.77, 95% confidence interval (CI)=0.64– 0.93, p=0.008; ΔLL, HR=0.79, 95% CI=0.67–0.95, p=0.01). Conclusion After AIS surgery, patients who had fusion to the lower lumbar vertebrae (L3 or L4) experienced a higher incidence and faster progression of degenerative changes in the SI joint. Low preoperative LL and inadequate correction of LL during the operation were also risk factors for SI joint degeneration.

Purpose: Improvement of left ventricular (LV) diastolic dysfunction (DD) is known to be a good prognostic factor in patients with heart failure with reduced ejection fraction (EF). In the present study, we investigated the predisposing risk factors affecting the reversibility of LV diastolic filling pattern (DFP) in patients with preserved EF. Materials and Methods: A total of 600 patients with pseudonormal LVDFP and preserved EF who underwent follow-up echocardiography were enrolled between 2011 and 2020. We compared their index and follow-up echocardiography findings and determined the predisposing risk factor affecting the reversibility of LVDFP. Results: Comparing the index and follow-up echocardiography findings showed that 379 (63%) patients had improved to normal or impaired relaxation LVDFP (improved group) and 221 (37%) patients had maintained or worsened LVDFP (unimproved group).The incidence of paroxysmal atrial fibrillation (PAF) was significantly higher in the unimproved group than in the improved group (4.7% vs. 9.5%, p=0.026). After adjustment for relevant clinical risk factors of diastolic dysfunction, PAF was determined to be an independent predisposing risk factor for the unimproved LVDFP (odds ratio: 2.10, 95% confidence interval: 1.06–4.15, p=0.033).Among the parameters of diastolic dysfunction in follow-up echocardiography, the left atrial volume index, mean E/A ratio, and E/e' were significantly improved in patients without PAF but remained in patients with PAF. Conclusion We identified that PAF was an independent predisposing risk factor of the unimproved LVDFP in patients with pseudonormal LVDFP and preserved EF. Therefore, early detection and management of PAF might be required in patients with LVDD and preserved EF to prevent adverse cardiovascular events.

Purpose: This study assessed the validity of the hospital standardized mortality ratio (HSMR) risk-adjusted model by comparing models that include clinical information and the current model based on administrative information in South Korea. Materials and Methods: The data of 53976 inpatients were analyzed. The current HSMR risk-adjusted model (Model 1) adjusts for sex, age, health coverage, emergency hospitalization status, main diagnosis, surgery status, and Charlson Comorbidity Index (CCI) using administrative data. As candidate variables, among clinical information, the American Society of Anesthesiologists score, Acute Physiology and Chronic Health Evaluation (APACHE) II, Simplified Acute Physiology Score (SAPS) 3, present on admission CCI, and cancer stage were collected. Surgery status, intensive care in the intensive care unit, and CCI were selected as proxy variables among administrative data. In-hospital death was defined as the dependent variable, and a logistic regression analysis was performed. The statistical performance of each model was compared using C-index values. Results: There was a strong correlation between variables in the administrative data and those in the medical records. The C-index of the existing model (Model 1) was 0.785; Model 2, which included all clinical data, had a higher C-index of 0.857. In Model 4, in which APACHE II and SAPS 3 were replaced with variables recorded in the administrative data from Model 2, the C-index further increased to 0.863. Conclusion The HSMR assessment model improved when clinical data were adjusted. Simultaneously, the validity of the evaluation method could be secured even if some of the clinical information was replaced with the information in the administrative data.

Background: s/Aims: Hepatocellular carcinoma (HCC) is the sixth most common cancer and second leading cause of cancer-related deaths in South Korea. This study evaluated the characteristics of Korean patients newly diagnosed with HCC in 2016-2018. Methods: Data from the Korean Primary Liver Cancer Registry (KPLCR), a representative database of patients newly diagnosed with HCC in South Korea, were analyzed. This study investigated 4,462 patients with HCC registered in the KPLCR in 2016-2018. Results: The median patient age was 63 years (interquartile range, 55-72). 79.7% of patients were male. Hepatitis B infection was the most common underlying liver disease (54.5%). The Barcelona Clinic Liver Cancer (BCLC) staging system classified patients as follows: stage 0 (14.9%), A (28.8%), B (7.5%), C (39.0%), and D (9.8%). The median overall survival was 3.72 years (95% confidence interval, 3.47-4.14), with 1-, 3-, and 5-year overall survival rates of 71.3%, 54.1%, and 44.3%, respectively. In 2016-2018, there was a significant shift toward BCLC stage 0-A and Child-Turcotte-Pugh liver function class A (P<0.05), although survival rates did not differ by diagnosis year. In the treatment group (n=4,389), the most common initial treatments were transarterial therapy (31.7%), surgical resection (24.9%), best supportive care (18.9%), and local ablation therapy (10.5%). Conclusions Between 2016 and 2018, HCC tended to be diagnosed at earlier stages, with better liver function in later years. However, since approximately half of the patients remained diagnosed at an advanced stage, more rigorous and optimized HCC screening strategies should be implemented.

Purpose: This study examined nationwide data regarding laparoscopic and robotic surgery for colorectal cancer (CRC) in Korea. Methods: Nationwide data concerning patients who underwent surgery for CRC from 2019 to 2023 were obtained from the Health Insurance Review and Assessment Service database. Results: From 2019 to 2023, a total of 109,573 patients with CRC underwent surgical resection in Korea. Among these, open, laparoscopic, and robotic surgery comprised 17.2%, 71.5%, and 11.3%, respectively. Open surgery decreased from 18.3% in 2019 to 15.2% in 2023, whereas robotic surgery increased from 10.3% in 2019 to 12.7% in 2023. Regarding rectal cancer, the rate of robotic surgery increased from 23.0% in 2019 to 28.2% in 2023, and the rate of minimally invasive surgery (MIS) increased from 86.9% in 2019 to 89.2% in 2023.Patients with National Health Insurance had significantly shorter lengths of hospital stay after surgery than those with medical aid for all surgical methods (p < 0.0001). With respect to hospital size, 74,282 CRC surgeries (67.8%) were performed in tertiary general hospitals and 33,050 (30.2%) in general hospitals. By the region, 47,140 cases (43.0%) were performed in Seoul, 19,961 (18.2%) in Gyeonggi, and 7,417 (6.8%) in Daegu. Ostomy was created in 16,222 CRC surgeries (14.8%). Conclusion The rate of MIS adoption for CRC in Korea has increased, reaching 84.7% in 2023. The rate of laparoscopic surgery exceeded 70% and has plateaued. In contrast, the rate of robotic surgery adoption has steadily increased, particularly for rectal cancer, where it surpassed 28% in 2023.

Objective: We aimed to investigate the associations of the triglyceride-glucose index, which measures insulin resistance, and the incidence of Parkinson’s disease. Methods: Our study used the Health Screening Cohort database of the National Health Insurance Service of South Korea (2002–2019). We included 310,021 participants who had no previous history of Parkinson’s disease and for whom more than 3 triglyceride-glucose index measurements were available. A diagnosis of Parkinson’s disease was determined via the International Classification of Diseases Tenth edition (G20) with a specific reimbursement code for rare intractable diseases and a history of prescriptions for anti-Parkinsonism drugs. Results: During a median of 9.64 years (interquartile range 8.72–10.53), 4,587 individuals (1.5%) had Parkinson’s disease. Based on a multivariable time-dependent Cox proportional hazards model, a per-unit increase in triglyceride-glucose index score was associated with a significantly increased risk of Parkinson’s disease (hazard ratio [HR]: 1.062; 95% confidence interval [CI] 1.007–1.119). In a sensitivity analysis, the triglyceride-glucose index was associated with the incidence of Parkinson’s disease in a non–diabetes mellitus cohort (HR: 1.093; 95% CI 1.025–1.165), but not in the diabetes mellitus cohort (HR: 0.990; 95% CI 0.902–1.087). In a restricted cubic spline analysis, the association between the triglyceride-glucose index and the incidence risk of Parkinson’s disease showed a nonlinear increasing (J-shaped) trend. Conclusion Our study demonstrated that higher triglyceride-glucose index scores were associated with the incidence of Parkinson’s disease in the general population, particularly in a nondiabetic mellitus cohort.

Background: Colorectal carcinomas (CRCs) with caudal-type homeobox 2 (CDX2) loss are recognized to pursue an aggressive behavior but tend to be accompanied by a high density of tumor-infiltrating lymphocytes (TILs). However, little is known about whether there is an interplay between CDX2 loss and TIL density in the survival of patients with CRC. Methods: Stage III CRC tissues were assessed for CDX2 loss using immunohistochemistry and analyzed for their densities of CD8 TILs in both intraepithelial (iTILs) and stromal areas using a machine learning-based analytic method. Results: CDX2 loss was significantly associated with a higher density of CD8 TILs in both intraepithelial and stromal areas. Both CDX2 loss and a high CD8 iTIL density were found to be prognostic parameters and showed hazard ratios of 2.314 (1.050–5.100) and 0.378 (0.175–0.817), respectively, for cancer-specific survival. A subset of CRCs with retained CDX2 expression and a high density of CD8 iTILs showed the best clinical outcome (hazard ratio of 0.138 [0.023–0.826]), whereas a subset with CDX2 loss and a high density of CD8 iTILs exhibited the worst clinical outcome (15.781 [3.939–63.230]). Conclusions Altogether, a high density of CD8 iTILs did not make a difference in the survival of patients with CRC with CDX2 loss. The combination of CDX2 expression and intraepithelial CD8 TIL density was an independent prognostic marker in adjuvant chemotherapy-treated patients with stage III CRC.

Background and Objectives: Obstructive sleep apnea (OSA) has been associated with an increased risk of cancer in various organs. OSA is also linked to chronic inflammation in the biliary tract and pancreas, a well-established risk factor for carcinogenesis in these organs. However, its relationship with biliary tract and pancreatic cancers remains unclear and has been rarely investigated. Therefore, we aimed to evaluate whether OSA serves as an independent risk factor for these malignancies by analyzing a nationwide healthcare claims database in South Korea. Methods: A retrospective cohort study was conducted using the Korean National Health Insurance Service (KNHIS) database. Adults aged ≥20 years who were newly diagnosed with OSA (ICD-10: G47.30) between 2007 and 2014 were identified and propensity score-matched (1:5) with controls based on age, sex, and comorbidities. Individuals with pre-existing cancer diagnoses were excluded. The primary endpoints were the incidence of overall cancer, biliary tract cancer (C23–C24), and pancreatic cancer (C25). Cox proportional hazards regression models were used to calculate hazard ratios (HRs), adjusting for demographic and clinical factors. Results: A total of 1,191,444 individuals were included, comprising 198,574 patients diagnosed with OSA and 992,870 matched controls. OSA was associated with an increased overall cancer incidence (HR, 1.132; 95% confidence interval [CI], 1.097–1.169); however, it was not significantly associated with pancreatic cancer (HR, 0.941; 95% CI, 0.823–1.072) or biliary tract cancer (HR, 0.931; 95% CI, 0.751–1.142). Subgroup analyses stratified by sex and age revealed no statistically significant associations across these groups. Conclusion Our findings do not support OSA as an independent risk factor for biliary tract or pancreatic cancers.

Although many wearable devices are used to assess sleep, their accuracy remains controversial. This study aimed to investigate the accuracy of the Actiwatch, a research-grade device, and the Fitbit, a consumer-grade device, against sleep diaries to assess sleep patterns. Methods: Twenty participants wore Fitbit and Actiwatch for two weeks and tracked their sleep patterns using sleep diaries. Total sleep time (TST), time-in-bed (TIB), sleep efficiency (SE), sleep onset latency (SOL), and wake after sleep onset (WASO) from the two devices and sleep diaries were analyzed using analysis of variance and Bland-Altman analysis. Results: The TIB measured by the sleep log, Fitbit, and Actiwatch were 420.9 minutes, 417.3 minutes, and 567.4 minutes, respectively. Compared to the sleep log, the Fitbit underestimated TST, TIB, and SE, with significant differences observed for TST (p<0.001) and SE (p<0.001), but not for TIB. The Actiwatch overestimated TIB (p<0.001) and TST (p=0.02) and underestimated SE (p<0.001) compared to the sleep log. The difference between the Fitbit and Actiwatch was significant for TST, TIB, and SE (all p<0.001). Conclusions: The Fitbit showed a smaller difference than the Actiwatch when compared with the sleep logs. The Fitbit could be used as a tool to assess sleep patterns in the clinic as well as in daily life.