Skull and intracranial metastases from hepatocellular carcinoma (HCC) have seldom been reported. A skull metastasis of HCC with a tumor bleeding resulting in spontaneous subdural hematoma (SDH) is extremely unusual. We report the first case of acute spontaneous SDH in a 69-year-old woman who presented with acute onset of headache, because of tumor bleeding caused by skull metastasis of HCC. A 69-year-old woman was referred to our hospital because of progressive headache, nausea, and vomiting for 3 days. Brain computed tomography (CT) performed in the emergency department (ED) revealed a left temporal SDH with a slight mass effect and a small left temporal bone erosion. Tri-phasic abdominal CT demonstrated a large right lobe liver tumor compatible with HCC. She experienced progressive deterioration of consciousness in the intensive care unit. Follow-up CT showed an enlargement of the SDH. An emergency craniotomy for hematoma evacuation and removal of skull tumor was performed. She regained consciousness and had no neurological deficits during the postoperative course. Pathological examination of the skull specimen indicated metastasis of a HCC. Patients with acute SDH without a history of head injury are rarely encountered in the ED. Metastatic carcinoma with bleeding should be included as a differential diagnosis for acute spontaneous SDH. Before an operation for SDH, the possibility of metastatic lesion of the skull should be considered in the surgical planning and the origin of malignancy should be sought.

The increasing prevalence of carbapenem-resistant Klebsiella pneumoniae (CRKP) infections is a global concern. Elderly patients have a diminished immune response and functional reserve, and are thus more vulnerable to bacterial infection. This study aimed to investigate the risk factors and outcomes in elderly patients with community-acquired CRKP infections. We performed a retrospective cohort study in a tertiary medical center between 1 January 2021, and 31 December 2021. All elderly patients who visited the emergency department during this period with culture-positive K. pneumoniae were enrolled, and their baseline demographics, laboratory profiles, management strategies, and outcomes were recorded and analyzed. We identified 528 elderly patients with K. pneumonia infection, and the proportion of patients with CRKP infection was 10.2% (54/528). Recent intensive care unit (ICU) admission and prior carbapenem use are independent risk factors for CRKP infection in elderly patients. Compared to patients with carbapenem-sensitive K. pneumoniae infection, those with CRKP infection had a significantly higher risk of adverse outcomes, including ICU care, respiratory failure, septic shock, and 90-day mortality. CRKP infection was also identified as an independent risk factor for 90-day mortality. Clinicians should be aware of the increasing prevalence of CRKP infections in elderly patients and judiciously choose appropriate antibiotics for these patients.

Background and Objectives: Chest X-ray (CXR) images are commonly used to diagnose respiratory and cardiovascular diseases. However, traditional manual interpretation is often subjective, time-consuming, and prone to errors, leading to inconsistent detection accuracy and poor generalization. In this paper, we present deep learning-based object detection methods for automatically identifying and annotating abnormal regions in CXR images. Methods: We developed and tested our models using disease-labeled CXR images and location-bounding boxes from E-Da Hospital. Given the prevalence of normal images over diseased ones in clinical settings, we created various training datasets and approaches to assess how different proportions of background images impact model performance. To address the issue of limited examples for certain diseases, we also investigated few-shot object detection techniques. We compared convolutional neural networks (CNNs) and Transformer-based models to determine the most effective architecture for medical image analysis. Results: The findings show that background image proportions greatly influenced model inference. Moreover, schemes incorporating binary classification consistently improved performance, and CNN-based models outperformed Transformer-based models across all scenarios. Conclusions: We have developed a more efficient and reliable system for the automated detection of disease labels and location bounding boxes in CXR images.

Convolutional neural networks (CNNs) have shown promise in accurately diagnosing coronavirus disease 2019 (COVID-19) and bacterial pneumonia using chest X-ray images. However, determining the optimal feature extraction approach is challenging. This study investigates the use of fusion-extracted features by deep networks to improve the accuracy of COVID-19 and bacterial pneumonia classification with chest X-ray radiography. A Fusion CNN method was developed using five different deep learning models after transferred learning to extract image features (Fusion CNN). The combined features were used to build a support vector machine (SVM) classifier with a RBF kernel. The performance of the model was evaluated using accuracy, Kappa values, recall rate, and precision scores. The Fusion CNN model achieved an accuracy and Kappa value of 0.994 and 0.991, with precision scores for normal, COVID-19, and bacterial groups of 0.991, 0.998, and 0.994, respectively. The results indicate that the Fusion CNN models with the SVM classifier provided reliable and accurate classification performance, with Kappa values no less than 0.990. Using a Fusion CNN approach could be a possible solution to enhance accuracy further. Therefore, the study demonstrates the potential of deep learning and fusion-extracted features for accurate COVID-19 and bacterial pneumonia classification with chest X-ray radiography.

Purpose: Klebsiella pneumoniae is an important causative pathogen of nosocomial infections, resulting in poor prognosis owing to its hypervirulence and antibiotic resistance. A simplified quicker version of the Pitt bacteremia score (PBS) (qPitt) for acute illness severity measurement was developed recently. The goal of this study was to explore the prognostic value of qPitt in patients with K. pneumoniae infection. Patients and Methods: Demographic information and management strategies were retrospectively collected from the records of all adult patients who visited the emergency department between January 1, 2021, and December 31, 2021, with culture-positive K. pneumoniae. The qPitt score was calculated based on: temperature < 36°C, systolic blood pressure ≤ 90 mmHg or vasopressor administration, respiratory rate ≥ 25 times/min or need of mechanical ventilation, altered mental status, and cardiac arrest event. The 30-day mortality prediction abilities of the qPitt were compared with the PBS, the sequential organ failure assessment (SOFA), and the quick sequential organ failure assessment (qSOFA) using receiver operating characteristic curves. Results: Data from 867 patients (57.8% men) with a mean age of 66.9 were compiled. The 30-day mortality rate of the enrolled patients was 13.4%, and the area under the curve (AUC) of the scoring systems were as follows: SOFA, 0.91 (95% confidence interval [CI]=0.89– 0.93), qPitt, 0.87 (95% CI=0.84– 0.89), PBS, 0.87 (95% CI=0.85– 0.89), and qSOFA, 0.73 (95% CI=0.70– 0.76). The AUC of qPitt was significantly higher than that of qSOFA (p< 0.01) and similar to that of PBS (p=0.65).The qPitt also demonstrated excellent mortality discrimination ability in non-bacteremic patients, AUC= 0.85 (95% CI=0.82– 0.88). Conclusion: The qPitt revealed excellent 30-day mortality prediction ability and also predicted mortality in non-bacteremic patients with K. pneumoniae infection. Clinicians can use this simplified scoring system to stratify patients earlier and initiate prompt treatment in high-risk patients.

Background: Klebsiella pneumoniae infection causes various diseases and leads to significant morbidity and mortality. The Pitt bacteremia score (PBS) is a well‐known prognostic predictor in patients with bacteremia. We aimed to investigate the prognostic role of the PBS in patients with nonbacteremic K. pneumoniae infections and compare its mortality discriminative ability with that of other risk scoring systems. Methods: Data were retrospectively collected from emergency department patients in E‐Da Hospital, Kaohsiung, Taiwan, within 2021. All adult patients (aged ≥ 20 years) during this period and diagnosed with K. pneumoniae infections were included. The baseline demographics, laboratory results, infection sources, and clinical outcomes of nonbacteremic patients were extracted, and the patients were further divided into low (< 4) and high (≥ 4) PBS groups for comparison. Results: A total of 863 patients with K. pneumoniae infection were identified, and 639 nonbacteremic patients were enrolled. There were similar demographics between the bacteremic and nonbacteremic groups. Regarding clinical outcomes in nonbacteremic patients, the high PBS group had significantly higher risk of septic shock (77.9% vs. 4.8%, p < 0.01), intensive care unit admission (71.3% vs. 8.2%, p < 0.01), respiratory failure (71.3% vs. 2.4%, p < 0.01), and 30‐day mortality (34.6% vs. 3.8%, p < 0.01). The area under the curve of the scoring systems regarding 30‐day mortality prediction ability was as follows: sequential organ failure assessment score 0.89 (95% confidence interval [CI] = 0.86–0.91), PBS 0.86 (95% CI = 0.83–0.88), quick sequential organ failure assessment score 0.71 (95% CI = 0.67–0.74), and systemic inflammatory response syndrome 0.62 (95% CI = 0.58–0.66). Conclusion: PBS correlated with adverse outcomes and good mortality prediction ability in patients with nonbacteremic K. pneumoniae infections.