Objective: We investigated the distribution of AI viruses in fecal samples from wild bird habitats (and nearby poultry-farm areas) surveyed between September and March from 2019 to 2025 and identified associated epidemiological risk factors. Methods: Samples were screened for influenza A (M, H5, H7) genes using real-time reverse transcription polymerase chain reaction (PCR), subjected to virus isolation in embryonated chicken eggs, and subtyped by PCR and sequencing. Host species were identified through DNA barcoding. Relative risks (RRs) with 95% confidence intervals were estimated for province, month, and waterfowl density. Results: Overall prevalence of HPAI and low pathogenic AI (LPAI) virus was 0.10% and 3.21%, respectively. HPAI virus was continuously isolated since 2020–2021, except 2019– 2020, while LPAI prevalence steadily increased (3.01%–4.35%). Twelve hemagglutinin (H1–H12) subtypes were identified in 1,722 isolates, and H3 (16.5%) was the most prevalent, followed by H5 (11.1%) and H7 (5.2%). LPAI H5N3 (55.7%) and H7N7 (75.5%) were the predominant H5 and H7 subtypes, respectively. Detection was higher in western coastal provinces, and higher mallard/spot-billed duck density and sampling in September– December were associated with increased risk. Conclusions and Relevance: Continued surveillance of migratory-bird habitats can provide early warning of HPAIV incursions and support targeted biosecurity measures in high-risk regions and seasons.

Background: The Act on Decisions on Life-Sustaining Treatment (LST) has been implemented in Korea since 2018, yet data on its application in neurocritical care units remain scarce. This study aimed to evaluate the clinical characteristics and outcomes of LST withdrawal or withholding in the neurocritical care unit. Methods: This study was a retrospective analysis conducted at a tertiary university hospital in Busan, South Korea. Among patients admitted to the neurocritical care unit between February 2018 and August 2023, those with documented decisions for LST withdrawal or withholding were enrolled. Demographic and clinical characteristics, underlying and combined conditions, reasons for LST decisions, measures taken, and time from LST withdrawal to death were extracted from medical records. Results: A total of 69 patients were included, with a median age of 67 years, and 38 (55%) were male. Cerebrovascular disease (62%) and traumatic brain injury (22%) were the most common underlying diagnoses. The primary reason for LST decisions was irreversible neurological damage (71%), followed by systemic complications (19%). Mechanical ventilation cessation (91%) and extubation (86%) were most frequently used measures for LST withdrawal. The median time from LST withdrawal to death was 22 minutes. Conclusions Our study demonstrates that LST decisions in the neurocritical care unit predominantly occur among patients with cerebrovascular disease or traumatic brain injury, mostly triggered by neurological deterioration. Most patients died shortly after withdrawal. These findings provide important insight into current LST withdrawal practices in neurocritical care and may assist clinical and ethical decision making in similar settings.

Background: The effectiveness of intravenous tissue plasminogen activator (IV tPA) in patients with large-vessel occlusion (LVO) receiving endovascular treatment (EVT) for acute ischemic stroke (AIS) has been questioned. We investigated IV tPA effectiveness in real-world AIS patients, including those with intracranial LVO receiving EVT. Methods: We identified patients with AIS who presented to hospital with National Institutes of Health Stroke Scale ≥4 within 8 hours of symptom onset from the institutional stroke registry. The association of IV tPA use with effectiveness and safety outcomes was analyzed in overall enrolled AIS patients; LVO patients; and patients treated with EVT. The effect of IV tPA was assessed using multiple logistic regression. Results: Among the 654 patients meeting study entry criteria, 238 (36.4%) received IV tPA and 416 (63.6%) did not. Multiple logistic regression analysis and shift analysis revealed IV tPA was associated with improved outcomes in overall enrolled AIS population, LVO, and EVT-treated subgroups. Among EVT-treated patients, IV tPA was associated with higher likelihood of ambulatory or better outcome (modified Rankin Scale 0–3) with odds ratio of 1.95 (P=0.03). Conclusions In this real-world study, IV tPA use was associated with improved outcomes for patients with AIS, including among LVO patients treated and not treated with EVT, in the contemporary mechanical thrombectomy era.

Background: Anastomotic leakage (AL) following esophagectomy represents a serious complication that often results in prolonged hospitalization and necessitates repeated interventions, including nothing-by-mouth (NPO) restriction, endoscopic vacuum therapy (EVT), or surgical repair. In this study, we evaluated the patterns and outcomes of AL treatment. Methods: We retrospectively reviewed the medical records of patients who underwent esophagectomy for esophageal cancer at a single center between 2003 and 2020. Of 3,096 examined cases, 181 patients (5.8%) with AL were included in the study: 114 patients (63%) with cervical anastomosis (CA) and 67 (37%) with intrathoracic anastomosis (TA). Results: The incidence of AL was 11.9% in the CA and 3.2% in the TA group (p<0.001).Among patients with CA who developed AL, 87 (76.3%) were managed with NPO, 15 (13.2%) with EVT, and 12 (10.5%) with surgical repair. Over 90% of patients with cervical AL resumed an oral diet by the time of discharge, regardless of treatment method. Among patients with TA and AL, 36 (53.7%) received NPO, 25 (37.7%) underwent EVT, and 6 (9%) required surgery. Of these, 34 patients who were managed with NPO and 19 with EVT could resume an oral diet. However, only 2 patients who underwent surgery resumed an oral diet, and 2 patients required additional EVT. Conclusion Although patients with CA displayed a higher incidence of AL, their rate of successful oral intake exceeded that of those with TA, regardless of treatment method.Among patients exhibiting AL with TA, EVT was more commonly employed than in CA cases, and it appears effective.

Background: This study assessed the comparative effectiveness of sextant and extended 12-core systematic biopsy within combined biopsy for the detection of prostate cancer. Methods: Patients who underwent combined biopsy targeting lesions with a Prostate Imaging Reporting and Data System (PI-RADS) score of 3–5 were assessed. Two specialists performed all combined cognitive biopsies. Both specialists performed target biopsies with five or more cores. One performed sextant systematic biopsies, and the other performed extended 12-core systematic biopsies. A total of 550 patients were analyzed. Results: Cases requiring systematic biopsy in combined biopsy exhibited a significant association with age ≥ 65 years (odds ratio [OR], 2.32; 95% confidence interval [CI], 1.25– 4.32; P = 0.008), PI-RADS score (OR, 2.32; 95% CI, 1.25–4.32; P = 0.008), and the number of systematic biopsy cores (OR, 3.69; 95% CI, 2.11–6.44; P < 0.001). In patients with an index lesion of PI-RADS 4, an extended 12-core systematic biopsy was required (target-negative/ systematic-positive or a greater Gleason score in the systematic biopsy than in the targeted biopsy) (P < 0.001). Conclusion During combined biopsy for prostate cancer in patients with PI-RADS 3 or 5, sextant systematic biopsy should be recommended over extended 12-core systematic biopsy when an effective targeted biopsy is performed.

Purpose: Since accurate lung cancer segmentation is required to determine the functional volume of a tumor in [ 18 F]FDG PET/CT, we propose a two-stage U-Net architecture to enhance the performance of lung cancer segmentation using [ 18 F]FDG PET/CT. Methods: The whole-body [ 18 F]FDG PET/CT scan data of 887 patients with lung cancer were retrospectively used for network training and evaluation. The ground-truth tumor volume of interest was drawn using the LifeX software. The dataset was randomly partitioned into training, validation, and test sets. Among the 887 PET/CT and VOI datasets, 730 were used to train the proposed models, 81 were used as the validation set, and the remaining 76 were used to evaluate the model. In Stage 1, the global U-net receives 3D PET/CT volume as input and extracts the preliminary tumor area, generating a 3D binary volume as output. In Stage 2, the regional U-net receives eight consecutive PET/CT slices around the slice selected by the Global U-net in Stage 1 and generates a 2D binary image as the output. Results: The proposed two-stage U-Net architecture outperformed the conventional one-stage 3D U-Net in primary lung cancer segmentation. The two-stage U-Net model successfully predicted the detailed margin of the tumors, which was determined by manually drawing spherical VOIs and applying an adaptive threshold. Quantitative analysis using the Dice similarity coefficient confirmed the advantages of the two-stage U-Net. Conclusion The proposed method will be useful for reducing the time and effort required for accurate lung cancer segmentation in [ 18 F]FDG PET/CT.

Frequent screening for thyroid cancer has been suggested as a probable explanation for the observed high risk of thyroid cancer in nuclear power plant (NPP) areas. We aimed to compare thyroid cancer screening rates of residents living near NPPs to those of the general population. This study utilized data from two national survey-based studies in 2016 and in 2014, respectively, for residents (n = 1,200) living in administrative districts within 5 km of NPP sites as the interest group, and the general population (n = 228,712) including distantliving residents (n = 19,100) in administrative districts within 30 km of NPP sites as reference groups. We observed an increase in screening rates in residents near NPPs, which may lead to a higher possibility of thyroid cancer detection. Therefore, further epidemiological studies investigating radiation-induced thyroid cancer risk among residents near NPPs should be carefully designed and interpreted considering possible detection bias.