Dongyang People's Hospital

Abstract Background: Sepsis is a leading cause of maternal mortality. Diagnosis in pregnancy is often delayed by non-specific presentation and legitimate caution regarding diagnostic imaging.This case illustrates a compounded diagnostic challenge: septic shock in a migrant patient, where unfamiliar, endemic comorbidities obscured the typical clinical picture. Case presentation: An 18-year-old primigravida at 27 weeks’ gestation presented with fever and mild respiratory symptoms, rapidly progressing to septic shock. Initial labs revealed severe microcytic anemia (Hb 59 g/L), marked hyperbilirubinemia, and leukocytosis. A low-dose abdominopelvic CT, crucial for source identification, revealed right-sided obstructive uropathy. This led to a diagnosis of obstructive pyelonephritis as the septic focus, managed with urgent ureteral stenting and antibiotics. Further investigation unmasked two underlying conditions contributing to her severe anemia: homozygous β-thalassemia intermedia and chronic Trichuris trichiura infection. The jaundice was attributed to sepsis-induced hepatocellular dysfunction. Following source control, transfusions, and supportive care, both maternal and fetal conditions stabilized. Conclusion: This case underscores that in pregnant patients with sepsis and atypical laboratory findings, timely imaging for source control is paramount. It also highlights the critical need to consider region-specific hereditary and parasitic diseases in the differential diagnosis for migrant populations, as these can significantly complicate the clinical presentation and management.

Abstract Background Caring behavior influence patients’ recovery and improving nursing quality.However,potential categories of caring behavior and the relationship between nurses’ caring behavior and moral sensitivity,remain unclear. Objective This study aimed to examine the current situation of nurses’ caring behavior,the correlation between nurses’ caring behavior and moral sensitivity,identify potential categories,and analyze the distribution characteristics of demographic variables in each subgroup. Methods The research was a descriptive, correlational study. Data were collected using Moral Sensitivity Questionnaire-Revised Version into Chinese,MSQR-CV(range: 9–54) and the Chinese version of Caring Behavior Inventory(CBI) (range: 24–144). A total of 387 nursing staff were seleted to be included in this research using a convenience sampling method.Latent Profile Analysis (LPA) was conducted to explore nurses’ caring behavior with 3 dimensions of the Chinese version of Caring Behavior Inventory as explicit variables.The influencing factors of different potential profiles of nurses’ caring behavior were analyzed using logistic regression analysis.Differences between profiles were analysed by MANOVA and ANOVAs as a follow-up. Results The LPA results showed that the three-profile model was the most suitable and supported the existence of three distinct QOL profiles: high(27.39%), moderate (34.11%) and low(8.4%). The relative entropy value was high (0.975), results pointed to a good profile solution and the three profiles differed significantly from one another. Conclusions The overall nurses’ caring behavior was at a moderate to high level.The classification of nurses could be predicted by factors, such as educational level, participation in caring training,grading of nurses and age.Furthermore, there was a positive correlation between the dimension of moral power and responsibility in moral sensitivity and caring behavior in nurses.On the contrary,lower dimension of moral burden,which was the negative dimension of moral sensitivity tended to higher caring behavior in nurses.Nursing managers can formulate targeted interventions according to the influencing factors of potential profiles to improve nurses’ caring behavior.

AIM: This study aimed to construct a risk stratification model for prostate cancer (PCa) ultrasound imaging data and machine learning algorithms, with the goal of providing an effective tool for early diagnosis, personalized treatment, and clinical decision-making. METHODS: A total of 211 histopathologically confirmed PCa patients were retrospectively enrolled and categorized into low-risk (n = 65), intermediate-risk (n = 55), and high-risk (n = 91) groups based on prostate-specific antigen levels, Gleason scores, and clinical T stage. From ultrasound images, 135 quantitative radiomic features—including morphological, texture, and edge descriptors—were extracted using the PyRadiomics toolkit. Feature dimensionality was reduced using the Pearson correlation coefficient (PCC), followed by recursive feature elimination (RFE) with 10-fold nested cross-validation to select the most informative features. Three machine learning algorithms—support vector machine (SVM), random forest (RF), and logistic regression (LR)—were trained and evaluated. Model performance was assessed using accuracy, sensitivity, specificity, and area under the curve (AUC). RESULTS: The RF model achieved the best performance in both training and test cohorts, with AUCs of 0.87 and 0.86, and accuracies of 90% and 88%, respectively. DeLong's test confirmed that RF significantly outperformed SVM (p = 0.016) and LR (p = 0.004) in AUC comparison. The RF model also demonstrated robust predictive ability across risk subgroups: in the high-risk group, it achieved an AUC of 0.89, accuracy of 89%, sensitivity of 88%, and specificity of 90%; in the intermediate- and low-risk groups, AUCs were 0.86 and 0.81, respectively. Feature importance analysis revealed that wavelet-transformed Gray Level Dependence Matrix (GLDM) texture features, particularly DependenceEntropy and DependenceVariance, were the most predictive, highlighting the value of intratumoral textural heterogeneity in risk classification. CONCLUSIONS: The RF-based ultrasound radiomics model enables accurate stratification of PCa risk, with remarkable performance in identifying high-risk patients.