Machine learning predictive model for aspiration screening in hospitalized patients with acute stroke.

Dysphagia is a fatal condition after acute stroke. We established machine learning (ML) models for screening aspiration in patients with acute stroke. This retrospective study enrolled patients with acute stroke admitted to a cerebrovascular specialty hospital between January 2016 and June 2022. A videofluoroscopic swallowing study (VFSS) confirmed aspiration. We evaluated the Gugging Swallowing Screen (GUSS), an early assessment tool for dysphagia, in all patients and compared its predictive value with ML models. Following ML algorithms were applied: regularized logistic regressions (ridge, lasso, and elastic net), random forest, extreme gradient boosting, support vector machines, k-nearest neighbors, and naïve Bayes. We finally analyzed data from 3408 patients, and 448 of them had aspiration on VFSS. The GUSS showed an area under the receiver operating characteristics curve (AUROC) of 0.79 (0.77-0.81). The ridge regression model was the best model among all ML models, with an AUROC of 0.81 (0.76-0.86), an F1 measure of 0.45. Regularized logistic regression models exhibited higher sensitivity (0.66-0.72) than the GUSS (0.64). Feature importance analyses revealed that the modified Rankin scale was the most important feature of ML performance. The proposed ML prediction models are valid and practical for screening aspiration in patients with acute stroke.

Recurrence and Mortality Risks in Patients with First Incident Acute Stroke or Myocardial Infarction: A Longitudinal Study Using the Korean National Health Insurance Service Database.

Background: We aimed to identify the long-term risk of recurrence and mortality in patients who experienced acute ischemic stroke (AIS), acute myocardial infarction (AMI), or acute hemorrhagic stroke (AHS) using a population-level database. Methods: This retrospective cohort study included adults aged ≥55 years diagnosed with AIS, AMI, and AHS in the National Health Insurance Service Database between 2004 and 2007. The target outcomes were secondary AIS, AMI, AHS, and all-cause mortality. Predetermined covariates, such as age, sex, socioeconomic status, hypertension, diabetes, and dyslipidemia, were adjusted. Results: We included 151,181, 49,077, and 41,636 patients in the AIS, AHS, and AMI groups, respectively. The AMI (adjusted hazard ratio [aHR], 0.318; 95% confidence interval [CI], 0.306−0.330; p < 0.001) and AHS (aHR, 0.489; 95% CI, 0.472−0.506; p < 0.001) groups had a significantly lower risk of developing secondary AIS than the AIS group. The risk of developing secondary AMI was significantly lower in the AMI (aHR, 0.388; 95% CI, 0.348−0.433; p < 0.001) and AHS (aHR, 0.711; 95% CI, 0.640−0.790; p < 0.001) groups than in the AIS group. Initial AHS was a decisive risk factor for secondary AHS (aHR, 8.546; 95% CI, 8.218−8.887; p < 0.001). The AMI (aHR, 1.436; 95% CI, 1.412−1.461; p < 0.001) and AHS (aHR, 1.328; 95% CI, 1.309−1.348; p < 0.001) groups were associated with a significantly higher risk of long-term mortality than the AIS group. Conclusion: Our results elucidated that initial AIS was a significant risk factor for recurrent AIS and AMI; initial AHS was a decisive risk factor for developing secondary AHS. Further, AMI and AHS were more closely related to long-term mortality than AIS.

Prognostic Value of Plasma Presepsin and Pneumonia Severity Index in Patients with Community-Acquired Pneumonia in the Emergency Department.

Background and Objectives: Presepsin (PSS) is an independent predictor for estimating disease severity and prognosis in septic patients. Few studies have reported the associations between plasma PSS and the severity and prognosis in patients with community-acquired pneumonia (CAP). We investigated whether a high plasma PSS level was associated with 30-day mortality in CAP patients. Materials and Methods: This retrospective single-center study was conducted in an emergency department. The PSS level was measured in 211 adult CAP patients admitted to the hospital and followed for up to 30 days. We recorded the pneumonia severity index (PSI) and the CURB-65 score. The primary outcome was death from any cause within 30 days. Results: The plasma PSS levels were significantly elevated in the high-risk group (PSI > 130) compared with the low- (PSI < 91) or moderate-risk groups (PSI 91−130). Forty-four patients (20.9%) died within 30 days of admission. Non-survivors had significantly higher plasma PSS levels than survivors among CAP patients: 1083 (697−1736) pg/mL vs. 385 (245−554) pg/mL (p < 0.001). The area under the curve (AUC) to predict 30-day mortality was highest for PSS (0.867), followed by procalcitonin (0.728) and lactate (0.616). The cutoff level of plasma PSS for 30-day mortality was >754 pg/mL. The combination of PSI and plasma PSS level improved the predictive ability for 30-day mortality (AUC = 0.892). Cox regression analysis showed that higher PSS levels (>754 pg/mL) and higher PSI (>126) were associated with 30-day mortality in CAP patients (hazard ratios of 19.472 and 6.375, respectively). Conclusion: Elevated plasma PSS is associated with severity and 30-day mortality in hospitalized CAP patients. Combining plasma PSS level and PSI could significantly improve the predictive ability of PSS for 30-day mortality.

Predictive Values of Procalcitonin and Presepsin for Acute Kidney Injury and 30-Day Hospital Mortality in Patients with COVID-19.

Background and Objectives: Acute kidney injury (AKI) is a common complication in patients with coronavirus disease 2019 (COVID-19). We investigated the values of procalcitonin (PCT) and presepsin (PSS) for predicting AKI and 30-day hospital mortality in patients with COVID-19. Materials and Methods: We retrospectively evaluated 151 patients with COVID-19 who were admitted to the hospital via the emergency department. The diagnosis of AKI was based on the Kidney Disease: Improving Global Outcomes clinical practice guidelines. Results: The median patient age was 77 years, and 86 patients (57%) were male. Fifty-six patients (37.1%) developed AKI, and 19 patients (12.6%) died within 30 days of hospital admission. PCT and PSS levels were significantly higher in patients with AKI and non-survivors. The cutoff values of PCT levels for predicting AKI and mortality were 2.26 ng/mL (sensitivity, 64.3%; specificity, 89.5%) and 2.67 ng/mL (sensitivity, 68.4%; specificity, 77.3%), respectively. The cutoff values of PSS levels for predicting AKI and mortality were 572 pg/mL (sensitivity, 66.0%; specificity, 69.1%) and 865 pg/mL (sensitivity, 84.6%; specificity, 76.0%), respectively. Conclusion: PCT and PSS are valuable biomarkers for predicting AKI and 30-day hospital mortality in patients with COVID-19.

The effects of socioeconomic and geographic factors on chronic phase long-term survival after stroke in South Korea.

The stroke incidence has increased rapidly in South Korea, calling for a national-wide system for long-term stroke management. We investigated the effects of socioeconomic status (SES) and geographic factors on chronic phase survival after stroke. We retrospectively enrolled 6994 patients who experienced a stroke event in 2009 from the Korean National Health Insurance database. We followed them up from 24 to 120 months after stroke onset. The endpoint was all-cause mortality. We defined SES using a medical-aid group and four groups divided by health insurance premium quartiles. Geographic factors were defined using Model 1 (capital, metropolitan, city, and county) and Model 2 (with or without university hospitals). The higher the insurance premium, the higher the survival rate tended to be (P < 0.001). The patient survival rate was highest in the capital city and lowest at the county level (P < 0.001). Regions with a university hospital(s) showed a higher survival rate (P = 0.006). Cox regression revealed that the medical-aid group was identified as an independent risk factor for chronic phase mortality. Further, NHIP level had a more significant effect than geographic factors on chronic stroke mortality. From these results, long-term nationwide efforts to reduce inter-regional as well as SES discrepancies affecting stroke management are needed.

Can Evoked Potential Changes during the Superficial Temporal Artery-Middle Cerebral Artery Bypass Surgery Predict Postoperative Improvement of Cerebral Perfusion and Functional Status?

BACKGROUND: We investigated evoked potential (EP) changes during superficial temporal artery to middle cerebral artery (STA-MCA) bypass surgery and their correlations with imaging and clinical findings postoperatively. METHODS: This retrospective study included patients who underwent STA-MCA bypass surgery due to ischemic stroke with large artery occlusion (MB group). Patients who underwent unruptured MCA aneurysm clipping were enrolled in the control group (MC group). Median and tibial somatosensory evoked potentials (SSEP), and motor evoked potentials recorded from the abductor pollicis brevis (APB-MEP) and abductor hallucis (AH-MEP) were measured intraoperatively. Modified Rankin scale (mRS) and perfusion-weighted imaging (PWI) related variables, i.e., mean transit time (MTT) and time to peak (TTP), were assessed. RESULTS: Δmedian SSEP, ΔAPB-MEP, and ΔAH-MEP were significantly higher in the MB group than in the MC group (p = 0.027, p = 0.006, and p = 0.015, respectively). APB-MEP and AH-MEP amplitudes were significantly increased at the final measurement (p = 0.010 and p < 0.001, respectively). The ΔTTP asymmetry index was moderately correlated with ΔAPB-MEP (r = 0.573, p = 0.005) and ΔAH-MEP (r = 0.617, p = 0.002). ΔAPB-MEP was moderately correlated with ΔMTT (r = 0.429, p = 0.047) and ΔmRS at 1 month (r = 0.514, p = 0.015). CONCLUSIONS: MEP improvement during STA-MCA bypass surgery was partially correlated with PWI and mRS and could reflect the recovery in cerebral perfusion.

Machine Learning-Based Three-Month Outcome Prediction in Acute Ischemic Stroke: A Single Cerebrovascular-Specialty Hospital Study in South Korea.

BACKGROUND: Functional outcomes after acute ischemic stroke are of great concern to patients and their families, as well as physicians and surgeons who make the clinical decisions. We developed machine learning (ML)-based functional outcome prediction models in acute ischemic stroke. METHODS: This retrospective study used a prospective cohort database. A total of 1066 patients with acute ischemic stroke between January 2019 and March 2021 were included. Variables such as demographic factors, stroke-related factors, laboratory findings, and comorbidities were utilized at the time of admission. Five ML algorithms were applied to predict a favorable functional outcome (modified Rankin Scale 0 or 1) at 3 months after stroke onset. RESULTS: Regularized logistic regression showed the best performance with an area under the receiver operating characteristic curve (AUC) of 0.86. Support vector machines represented the second-highest AUC of 0.85 with the highest F1-score of 0.86, and finally, all ML models applied achieved an AUC > 0.8. The National Institute of Health Stroke Scale at admission and age were consistently the top two important variables for generalized logistic regression, random forest, and extreme gradient boosting models. CONCLUSIONS: ML-based functional outcome prediction models for acute ischemic stroke were validated and proven to be readily applicable and useful.

Usefulness of Intraoperative Neurophysiological Monitoring During the Clipping of Unruptured Intracranial Aneurysm: Diagnostic Efficacy and Detailed Protocol.

Background: Intraoperative neurophysiological monitoring (IONM) has been widely applied in brain vascular surgeries to reduce postoperative neurologic deficit (PND). This study aimed to investigate the effect of IONM during clipping of unruptured intracranial aneurysms (UIAs). Methods: Between January 2013 and August 2020, we enrolled 193 patients with 202 UIAs in the N group (clipping without IONM) and 319 patients with 343 UIAs in the M group (clipping with IONM). Patients in the M group were intraoperatively monitored for motor evoked potentials (MEPs) and somatosensory evoked potentials (SSEPs). Irreversible evoked potential (EP) change was defined as EP deterioration that did not recover until surgery completion. Sustained PND was defined as neurological symptoms lasting for more than one postoperative month. Results: Ten (3.1%) and 13 (6.7%) in the M and N groups, respectively, presented with PND. Compared with the N group, the M group had significantly lower occurrence rates of sustained PND [odds ratio (OR) = 0.36, 95% confidence interval (CI) = 0.13-0.98, p = 0.04], ischemic complications (OR = 0.39, 95% CI = 0.15-0.98, p = 0.04), and radiologic complications (OR = 0.40, 95% CI = 0.19-0.82, p = 0.01). Temporary clipping was an independent risk factor for ischemic complications (ICs) in the total patient group (OR = 6.18, 95% CI = 1.75-21.83, p = 0.005), but not in the M group (OR = 5.53, 95% CI = 0.76-41.92, p = 0.09). Regarding PND prediction, considering any EP changes (MEP and/or SSEP) showed the best diagnostic efficiency with a sensitivity of 0.900, specificity of 0.940, positive predictive value of 0.321, negative predictive value (NPV) of 0.997, and a negative likelihood ratio (LR) of 0.11. Conclusion: IONM application during UIA clipping can reduce PND and radiological complications. The diagnostic effectiveness of IONM, specifically the NPV and LR negative values, was optimal upon consideration of changes in any EP modality.

Expression of VEGF, HGF, IL-6, IL-8, MMP-9, Telomerase in Peripheral Blood of Patients with Head and Neck Squamous Cell Carcinoma.

OBJECTIVES: This study investigated the telomerase expression in peripheral blood mononuclear cells (PBMCs) and the relationship between the serum level of several soluble factors such as vascular endothelial growth factor (VEGF), hepatocyte growth factor, interleukin (IL)-6, IL-8, and matrix metallopeptidase-9 and the clinicopathological features of patients with head and neck squamous cell carcinoma (HNSCC). METHODS: Peripheral blood samples were collected from 50 HNSCC patients and 15 normal controls. The telomerase activity in the PBMCs was measured by Telomere Repeat Amplification Protocols. The serum levels of the soluble factors were analyzed by enzyme-linked immunosorbent assay. RESULTS: The expression of telomerase in the PBMCs of HNSCC patients was significantly correlated with the N and American Joint Committee on Cancer (AJCC) stages. The serum VEGF level was significantly higher in the patients with an advanced T stage, N stage and AJCC stage. Serum VEGF was significantly related with the expression of telomerase in the PBMCs. The telomerase expression and the VEGF expression were shown to be independent factors associated with poor survival. CONCLUSION: The telomerase expression in the PBMCs and the serum VEGF level of HNSCC patients were significantly correlated with the N stage, the AJCC stage and the prognosis.