OBJECTIVES: We propose a novel deep learning segmentation algorithm (DSRF) based on multi-scale supervision and residual feedback strategy for precise segmentation of the optic chiasm and optic nerves in CT images of nasopharyngeal carcinoma (NPC) patients. METHODS: We collected 212 NPC CT images and their ground truth labels from SegRap2023, StructSeg2019 and HaN-Seg2023 datasets. Based on a hybrid pooling strategy, we designed a decoder (HPS) to reduce small organ feature loss during pooling in convolutional neural networks. This decoder uses adaptive and average pooling to refine high-level semantic features, which are integrated with primary semantic features to enable network learning of finer feature details. We employed multi-scale deep supervision layers to learn rich multi-scale and multi-level semantic features under deep supervision, thereby enhancing boundary identification of the optic chiasm and optic nerves. A residual feedback module that enables multiple iterations of the network was designed for contrast enhancement of the optic chiasm and optic nerves in CT images by utilizing information from fuzzy boundaries and easily confused regions to iteratively refine segmentation results under supervision. The entire segmentation framework was optimized with the loss from each iteration to enhance segmentation accuracy and boundary clarity. Ablation experiments and comparative experiments were conducted to evaluate the effectiveness of each component and the performance of the proposed model. RESULTS: The DSRF algorithm could effectively enhance feature representation of small organs to achieve accurate segmentation of the optic chiasm and optic nerves with an average DSC of 0.837 and an ASSD of 0.351. Ablation experiments further verified the contributions of each component in the DSRF method. CONCLUSIONS The proposed deep learning segmentation algorithm can effectively enhance feature representation to achieve accurate segmentation of the optic chiasm and optic nerves in CT images of NPC.
Humans ; Tomography, X-Ray Computed/methods* ; Optic Chiasm/diagnostic imaging* ; Optic Nerve/diagnostic imaging* ; Algorithms ; Nasopharyngeal Carcinoma ; Deep Learning ; Nasopharyngeal Neoplasms/diagnostic imaging* ; Neural Networks, Computer ; Image Processing, Computer-Assisted/methods*

Objective To establish a predictive model for survival outcomes of glioma patients based on both brain radiomics features from preoperative MRI multi-sequence images and clinical features.Methods We retrospectively analyzed the MRI images and clinical data of 388 glioma patients and extracted the radiomics features from the peritumoral edema zone,tumor core,and whole tumor on T1,T2,and T1-weighted contrast-enhanced(T1CE)and fluid attenuated inversion recovery(FLAIR)sequences.The cases were divided into a training set(271 cases)and a test set(117 cases).Random survival forest algorithms were used to select the radiomics features associated with overall survival(OS)in the training set to construct a radiomic score(Rad-score),based on which the patients were classified into high-and low-risk groups for Kaplan-Meier survival analysis.Cox proportional hazard regression models for the 3 different tumor zones were constructed,and their performance for predicting 1-and 3-year survival rates was evaluated using 5-fold cross-validation and AUC analysis followed by external validation using data from another 10 glioma patients.The best-performing model was used for constructing a nomogram for survival predictions.Results Five radiomics features from the tumor core,7 from the peritumoral edema zone,and 5 from the whole tumor were selected.In both the training and test sets,the high-and low-risk groups had significantly different OS(P<0.05),and age,IDH status and Rad-score were independent factors affecting OS.The combined model showed better performance than the Rad-score model with AUCs for 1-year and 3-year survival prediction of 0.750 and 0.778 in the training set,0.764 and 0.800 in the test set,and 0.938 and 0.917 in external validation,respectively.Conclusion The predictive model combining preoperative multi-modal MRI radiomics features and clinical features can effectively predict survival outcomes of glioma patients.

Objective To establish a predictive model for survival outcomes of glioma patients based on both brain radiomics features from preoperative MRI multi-sequence images and clinical features.Methods We retrospectively analyzed the MRI images and clinical data of 388 glioma patients and extracted the radiomics features from the peritumoral edema zone,tumor core,and whole tumor on T1,T2,and T1-weighted contrast-enhanced(T1CE)and fluid attenuated inversion recovery(FLAIR)sequences.The cases were divided into a training set(271 cases)and a test set(117 cases).Random survival forest algorithms were used to select the radiomics features associated with overall survival(OS)in the training set to construct a radiomic score(Rad-score),based on which the patients were classified into high-and low-risk groups for Kaplan-Meier survival analysis.Cox proportional hazard regression models for the 3 different tumor zones were constructed,and their performance for predicting 1-and 3-year survival rates was evaluated using 5-fold cross-validation and AUC analysis followed by external validation using data from another 10 glioma patients.The best-performing model was used for constructing a nomogram for survival predictions.Results Five radiomics features from the tumor core,7 from the peritumoral edema zone,and 5 from the whole tumor were selected.In both the training and test sets,the high-and low-risk groups had significantly different OS(P<0.05),and age,IDH status and Rad-score were independent factors affecting OS.The combined model showed better performance than the Rad-score model with AUCs for 1-year and 3-year survival prediction of 0.750 and 0.778 in the training set,0.764 and 0.800 in the test set,and 0.938 and 0.917 in external validation,respectively.Conclusion The predictive model combining preoperative multi-modal MRI radiomics features and clinical features can effectively predict survival outcomes of glioma patients.

Objective:To investigate the potential source of infection for a cluster outbreak of human brucellosis in Yantai City, Shandong Province.Methods:The information of a human brucellosis cluster outbreak case in Yantai City, Shandong Province in 2022 was collected, the strains were isolated and cultured, and DNA was extracted. BCSP31-PCR was used for species identification, and AMOS-PCR was used for species type identification. Multiple locus variable-number tandem-repeat analysis (MLVA)-16 was used for clustering analysis, and the results were compared with the public database MLVAbank and the monitoring data of Brucella in Shandong Province in 2022. At the same time, whole genome single nucleotide polymorphism (wgSNP) typing was used to analyze the 53 Brucella strains that had completed whole genome sequencing in Shandong Province in 2022, and the wgSNP phylogenetic tree was constructed. Results:According to BCSP31-PCR and AMOS-PCR identification, the three strains related to the cluster outbreak of brucellosis in Yantai City, Shandong Province in 2022 were all Brucella melitensis biotype. The results of MLVA-16 typing showed that the MLVA-16 typing of the three isolated strains was completely consistent, with 16 tandem repeat loci of 1-5-3-13-2-2-3-2-4-41-8-4-4-3-6-5, belonging to the Eastern Mediterranean clade. Compared with MLVAbank, the MLVA-16 typing of two strains isolated from Kazakhstan was consistent with the results of this study. Compared with the monitoring data of Brucella in Shandong Province in 2022, it was found that the MLVA-16 typing of 11 isolated strains was consistent with the results of this study, which were isolated from Zaozhuang, Linyi, Taian, Yantai, and Weifang cities, respectively. The results of wgSNP typing showed that the distance between the 11 strains and the strains of the current outbreak was less than 7 single nucleotide polymorphisms, and the strains were isolated from Taian, Zibo, Linyi, Binzhou, Jinan, Jining, Yantai and Weihai cities, respectively. Conclusion:After tracing the source of a human brucellosis cluster outbreak in Yantai City, Shandong Province in 2022, it is speculated that the strains of Brucella melitensis isolated from Linyi, Taian and Yantai cities are closely related, indicating that sheep in these areas have homology.

Objective:To deeply explore the decision-making experience of patients participating in clinical trials of neoadjuvant immunotherapy for lung cancer.Methods:Using the descriptive and qualitative research, 15 lung cancer patients who participated in clinical trials of neoadjuvant immunotherapy at the Cancer Hospital of the Chinese Academy of Medical Sciences were selected by purposive sampling from April 2021 to August 2022 for semi-structured in-depth interviews. Content analysis method was used for data analysis, summarization, and induction.Results:Three themes were extracted, namely decision-making information dilemma (insufficient or overloaded information, difficulty in understanding professional information, urgent need for decision-making information assistance), complex emotional experience (negative emotional experience, positive emotional experience), and hope for multi-party support (expecting psychological communication, hoping for family understanding, and longing for social recognition) .Conclusions:The decision-making experience of patients in clinical trials of neoadjuvant immunotherapy for lung cancer were summarized and described, which can help strengthen the understanding of the research team and medical and nursing staff on the live experience of such patients when making decisions, provide targeted decision support strategies, and promote good informed consent of patients.

Objective:To assess the clinical value of nomogram based on high resolution magnetic resonance vessel wall imaging (HR-VWI) features in differentiating moyamoya disease (MMD) from atherosclerotic moyamoya syndrome (A-MMS).Methods:Eighty-four patients with digital subtraction angiography (DSA)-confirmed MMD and 73 patients with DSA-confirmed A-MMS were enrolled from Department of Medical Imaging, Affiliated Hospital of Jining Medical University from June 2020 to November 2023. All patients underwent HR-VWI. A retrospective analysis was performed on their imaging data. Univariate analysis was used to compare the differences in imaging characteristics between the two groups. Multivariate Logistic regression analysis was used to screen independent influencing factors for differentiating MMD from A-MMS and a nomogram was constructed accordingly. Receiver operating characteristic (ROC) curve and calibration curve were used to evaluate the validity and calibration of the nomogram.Results:Univariate analysis showed that age, pattern of vessel wall thickening, maximum vessel wall thickness, enhancement degree of vessel wall, vessel external diameter, ipsilateral anterior cerebral artery involvement and dot sign were statistically different between the two groups ( P<0.05). Multivariate Logistic regression analysis showed that age ( OR=6.990, 95% CI: 2.340-20.360, P<0.001), pattern of vessel wall thickening ( OR=0.066, 95% CI: 0.014-0.307, P<0.001), vessel external diameter ( OR=5.224, 95% CI: 1.672-16.324, P=0.005), ipsilateral anterior cerebral artery involvement ( OR=0.160, 95% CI: 0.038-0.679, P=0.013) and dot sign ( OR=0.081, 95% CI: 0.018-0.364, P=0.001) were independent influencing factors for differentiating MMD from A-MMS. ROC curve showed that area under the curve (AUC) of this nomogram was 0.884 (95% CI: 0.821-0.947, P<0.001), and the calibration curve showed a good fit between the predicted probability and actual probability. Conclusion:Nomogram based on HR-VWI features can effectively differentiate MMD from A-MMS.

Background: Data on the efficacy and incidence of adverse effects associated with dexmedetomidine (DEX) as a local anesthetic adjuvant for patient-controlled epidural analgesia (PCEA) are inconclusive. This meta-analysis assessed the efficacy and risks of DEX for PCEA using opioids as a reference. Methods: Two researchers independently searched PubMed, Embase, Cochrane Library, and China Biology Medicine for randomized controlled trials comparing DEX and opioids as local anesthetic adjuvants in PCEA. Results: In total, 636 patients from seven studies were included in this meta-analysis. Postoperative patients who received DEX had lower visual analog scale (VAS) scores than those who received opioids at 4–8 h (mean difference [MD]: 0.61, 95% CI [0.45, 0.76], P < 0.001, I2 = 0%), 12 h (MD: 0.85, 95% CI [0.61, 1.09], P < 0.001, I2 = 0%), 24 h (MD: 0.59, 95% CI [0.06, 1.12], P = 0.030, I2 = 82%), and 48 h (MD: 0.54, 95% CI [0.05, 1.02], P = 0.030, I2 = 91%). Additionally, patients who received DEX had a lower incidence of itching (odds ratio [OR]: 2.86, 95% CI [1.18, 6.95], P = 0.020, I2 = 0%) and nausea and vomiting (OR: 6.83, 95% CI [3.63, 12.84], P < 0.001, I2 = 24%). In labor analgesia, no significant differences in neonatal (pH and PaO2 of cord blood, fetal heart rate) or maternal outcomes (duration of labor stage, mode of delivery) were found between the DEX and opioid groups. Conclusions Compared with opioids, using DEX as a local anesthetic adjuvant in PCEA improved postoperative analgesia and reduced the incidence of itching and nausea and vomiting without increasing the incidence of adverse events.

Objective: To investigate the relationship between insulin resistance and the serum level of mesencephalic astrocyte⁃derived neurotrophic factor (MANF) in patients with polycystic ovary syndrome (PCOS) . Methods: There were 120 patients with PCOS and 40 healthy women recruited as the experimental group and the control group respectively. The serum levels of sex hormones , fasting blood glucose ( FBG) , fasting insulin ( FINS) , and MANF in the experimental group and the control group were compared. The serum MANF level of patients with or without insulin resistance in the experimental group was also compared and the relationship between homeostasis model assessment for insulin resistance (HOMA⁃IR) and the serum MANF level was analyzed. Results: Compared to the control group , the serum levels of luteinizing hormone , testosterone , androstenedione , dehydroepiandrosterone sulfate , FBG , FINS , and HOMA⁃IR in the experimental group were significantly higher (P < 0. 05) ,whereas the serum levels of follicle stimulating hormone , sex hormone binding globulin , and MANF were significantly lower (P <0. 05) . Compared to patients without insulin resistance (59 cases) , the serum MANF level of patients with insulin resistance in the experimental group (61 cases) was significantly lower (P < 0. 05) , and there was a negative correlation between the serum MANF level and HOMA⁃IR of PCOS patients in the experimental group ( P < 0. 05 ) . Conclusion MANF possibly plays an important role in the occurrence and progression of insulin resistance in PCOS patients ,which will be able to provide new approaches to the diagnosis and treatment of PCOS.

Objective:To analyze the risk factors of acute kidney injury (AKI) in hip fracture patients with serious underlying diseases and establish a prediction nomogram.Methods:Clinical information of hip fracture patients admitted to the intensive care unit (ICU) of Beth Israel Deaconess Medical Center (BIDMC) was analyzed using the Medical Information Mart for Intensive Care (MIMIC)-IV. Patient comorbidities, disease scores, vital signs and laboratory tests, surgical modalities, invasive procedures, and drug use were recorded. According to the diagnostic criteria of AKI in the Kidney Disease Improving Global Outcome (KDIGO) guideline, the enrolled patients were randomly divided into training set and validation set. Based on logistic regression analysis, least absolute shrinkage and selection operator (LASSO) logistic regression algorithm was used to analyze the risk factors of AKI after admission, and the corresponding prediction model was calculated.Results:A total of 474 patients were enrolled, including 331 in the training set and 143 in the validation set. According to the diagnostic criteria of AKI of KDIGO guidelines, the patients were divided into AKI group (159 cases) and non-AKI group (172 cases). Univariate analysis showed that age ( t=2.61, P=0.009), coronary heart disease (χ 2=2.08, P=0.038), heart failure (χ 2=2.60, P=0.009), hemoglobin ( t=1.89, P=0.059), platelets ( t=1.81, P=0.070), urea nitrogen ( t=2.83, P=0.005), blood creatinine ( t=3.65, P<0.001), blood sodium ( t=2.55, P=0.011), blood glucose ( t=2.52, P=0.012), anion gap ( t=3.44, P=0.001), diastolic blood pressure ( t=2.72, P=0.007), mean arterial pressure ( t=2.16, P=0.031), SOFA score ( t=3.69, P<0.001), simplified acute physiological function score II (SAPSII) score ( t=2.95, P=0.003), as well as furosemide (χ 2=2.03, P=0.042), vancomycin (χ 2=1.70, P=0.089), vasoactive medications (χ 2=3.74, P<0.001) and use of invasive mechanical ventilation (χ 2=4.81, P<0.001) were risk factors associated with the development of AKI in hip fracture patients. Multivariate logistic regression analysis showed that age ( OR=1.03, P<0.001), coronary heart disease ( OR=2.05, P=0.069), hemoglobin ( OR=0.88, P=0.050), blood creatinine ( OR=1.37, P=0.009), blood sodium ( OR=1.07, P=0.026), anion gap ( OR=1.09, P=0.028) and vasoactive medications ( OR=3.83, P=0.018) and the use of invasive mechanical ventilation ( OR=6.56, P<0.001) were independent predictors of the development of AKI in hip fracture patients with serious underlying diseases. The area under the curve of the nomogram prediction model constructed by the above 8 predictors was 0.789, and the calibration curve of the nomogram was close to the ideal diagonal. Decision curve analysis showed that the net benefit of the model was significant. Conclusion:The incidence of AKI is high in hip fracture patients with serious underlying diseases. Age, coronary heart disease, hemoglobin, serum creatinine, serum sodium, anion gap, vasoactive drugs, and invasive mechanical ventilation can predict the occurrence of AKI to a certain extent. Combined with the risk factors, the construction of the corresponding prediction model can predict and manage the diagnosis and treatment of AKI in patients with hip fracture complicated with severe underlying diseases.