Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Guided by the concept of quality by design(QbD), this study optimizes the calcination and quenching process of calcined Magnetitum and establishes the XRD characteristic spectra of calcined Magnetitum, providing a scientific basis for the formulation of quality standards. Based on the processing methods and quality requirements of Magnetitum in the Chinese Pharmacopoeia, the critical process parameters(CPPs) identified were calcination temperature, calcination time, particle size, laying thickness, and the number of vinegar quenching cycles. The critical quality attributes(CQAs) included Fe mass fraction, Fe~(2+) dissolution, and surface color. The weight coefficients were determined by combining Analytic Hierarchy Process(AHP) and the criteria importance though intercrieria correlation(CRITIC) method, and the calcination process was optimized using orthogonal experimentation. Surface color was selected as a CQA, and based on the principle of color value, the surface color of calcined Magnetitum was objectively quantified. The vinegar quenching process was then optimized to determine the best processing conditions. X-ray diffraction(XRD) was used to establish the characteristic spectra of calcined Magnetitum, and methods such as similarity evaluation, cluster analysis, and orthogonal partial least squares-discriminant analysis(OPLS-DA) were used to evaluate the quality of the spectra. The optimized calcined Magnetitum preparation process was found to be calcination at 750 ℃ for 1 h, with a laying thickness of 4 cm, a particle size of 0.4-0.8 cm, and one vinegar quenching cycle(Magnetitum-vinegar ratio 10∶3), which was stable and feasible. The XRD characteristic spectra analysis method, featuring 9 common peaks as fingerprint information, was established. The average correlation coefficient ranged from 0.839 5-0.988 1, and the average angle cosine ranged from 0.914 4 to 0.995 6, indicating good similarity. Cluster analysis results showed that Magnetitum and calcined Magnetitum could be grouped together, with similar compositions. OPLS-DA discriminant analysis identified three key characteristic peaks, with Fe_2O_3 being the distinguishing component between the two. The final optimized processing method is stable and feasible, and the XRD characteristic spectra of calcined Magnetitum was initially established, providing a reference for subsequent quality control and the formulation of quality standards for calcined Magnetitum.
X-Ray Diffraction/methods* ; Drugs, Chinese Herbal/chemistry* ; Quality Control ; Particle Size

Retrieving keyframes most relevant to text from small intestine videos with given labels can efficiently and accurately locate pathological regions. However, training directly on raw video data is extremely slow, while learning visual representations from image-text datasets leads to computational inconsistency. To tackle this challenge, a small bowel video keyframe retrieval based on multi-modal contrastive learning (KRCL) is proposed. This framework fully utilizes textual information from video category labels to learn video features closely related to text, while modeling temporal information within a pretrained image-text model. It transfers knowledge learned from image-text multimodal models to the video domain, enabling interaction among medical videos, images, and text data. Experimental results on the hyper-spectral and Kvasir dataset for gastrointestinal disease detection (Hyper-Kvasir) and the Microsoft Research video-to-text (MSR-VTT) retrieval dataset demonstrate the effectiveness and robustness of KRCL, with the proposed method achieving state-of-the-art performance across nearly all evaluation metrics.
Humans ; Video Recording ; Intestine, Small/diagnostic imaging* ; Machine Learning ; Image Processing, Computer-Assisted/methods* ; Algorithms

OBJECTIVE: To investigate the clinical efficacy and safety of ultrasound-guided closed reduction and internal fixation using Kirschner wire for the treatment of olecranon fractures of the ulna in children. METHODS: Between January 2019 and January 2021, 13 children with olecranon fracture were treated with ultrasound-guided closed reduction and percutaneous Kirschner wire internal fixation, including 10 males and 3 females. The age ranged from 3 to 14 years old. Children with ulnar olecranon fractures were evaluated using the Gicquel scoring system. The clinical evaluation encompassed postoperative pain, functional status, and range of motion, with a maximum score of 15 points. The radiological assessment contributed an additional 4 points. A cumulative score of more than 18 scores was classified as excellent, more than 17 scores as good, more than16 scores as fair, and less than 16 scores as poor. Clinical assessment:A score of 14 indicates excellent performance, a score of 13 reflects good performance, a score of 12 denotes fair performance, and a score of less than 11 signifies poor performance. RESULTS: A total of 13 patients were followed up, with a duration ranging from 6 to 12 months. According to the Gicquel scoring criteria, the comprehensive evaluation of clinical and radiographic findings yielded 10 excellent and 3 good outcomes. Evaluation based solely on clinical findings resulted in 13 excellent outcomes. CONCLUSION Ultrasound-guided percutaneous cross Kirschner wire fixation for children's olecranon fracture has the advantages of less trauma, rapid recovery, less fluoroscopy, and good recovery of elbow function. The clinical effect is satisfactory.
Humans ; Child ; Male ; Female ; Fracture Fixation, Internal/instrumentation* ; Ulna Fractures/physiopathology* ; Bone Wires ; Child, Preschool ; Adolescent ; Olecranon Process/surgery* ; Ultrasonography ; Closed Fracture Reduction/methods* ; Olecranon Fracture

The relationship between hyperuricemia (HUA) and erectile dysfunction (ED) remains inadequately understood. Given that HUA is often associated with various metabolic disorders, this study aims to explore the multivariate linear impacts of metabolic parameters on erectile function in ED patients with HUA. A cross-sectional analysis was conducted involving 514 ED patients with HUA in the Department of Andrology, Jiangsu Province Hospital of Chinese Medicine (Nanjing, China), aged 18 to 60 years. General demographic information, medical history, and laboratory results were collected to assess metabolic disturbances. Sexual function was evaluated using the 5-item version of the International Index of Erectile Function (IIEF-5) questionnaire. Based on univariate analysis, variables associated with IIEF-5 scores were identified, and the correlations between them were evaluated. The effects of these variables on IIEF-5 scores were further explored by multiple linear regression models. Fasting plasma glucose ( β = -0.628, P < 0.001), uric acid ( β = -0.552, P < 0.001), triglycerides ( β = -0.088, P = 0.047), low-density lipoprotein cholesterol ( β = -0.164, P = 0.027), glycated hemoglobin (HbA1c; β = -0.562, P = 0.012), and smoking history ( β = -0.074, P = 0.037) exhibited significant negative impacts on erectile function. The coefficient of determination ( R ²) for the model was 0.239, and the adjusted R ² was 0.230, indicating overall statistical significance ( F -statistic = 26.52, P < 0.001). Metabolic parameters play a crucial role in the development of ED. Maintaining normal metabolic indices may aid in the prevention and improvement of erectile function in ED patients with HUA.
Humans ; Male ; Erectile Dysfunction/metabolism* ; Hyperuricemia/metabolism* ; Adult ; Middle Aged ; Cross-Sectional Studies ; Glycated Hemoglobin/metabolism* ; Blood Glucose/metabolism* ; Uric Acid/blood* ; Young Adult ; Triglycerides/blood* ; Adolescent ; Cholesterol, LDL/blood* ; Penile Erection/physiology* ; Surveys and Questionnaires

OBJECTIVE: To explore the application of targeted mRNA sequencing in fusion gene diagnosis of hematologic diseases. METHODS: Bone marrow or peripheral blood samples of 105 patients with abnormally elevated eosinophil proportions and 291 acute leukemia patients from January 2015 to June 2023 in the First Affiliated Hospital of Soochow University were analyzed and gene structural variants were detected by targeted mRNA sequencing. RESULTS: Among 105 patients with abnormally elevated eosinophil proportions, 6 cases were detected with gene structural variants, among which fusion gene testing results in 5 cases could serve as diagnostic indicators for myeloid neoplasms with eosinophilia. In addition, a IL3∷ETV6 fusion gene was detected in one patient with chronic eosinophilic leukemia, not otherwise specified. Among 119 patients with acute myeloid leukemia (AML), 38 cases were detected structural variants by targeted mRNA sequencing, accounting for 31.9%, which was significantly higher than 20.2% (24/119) detected by multiple quantitative PCR (P < 0.05). We also found one patient with AML had both NUP98∷PRRX2 and KCTD5∷JAK2 fusion genes. A total of 104 patients were detected structural variants by targeted mRNA sequencing in 172 cases with acute B-lymphoblastic leukemia who were tested negative by multiple quantitative PCR, with a detection rate of 60.5% (102/172). CONCLUSION Targeted mRNA sequencing can effectively detect fusion gene and has potential clinical application value in diagnosis and classificatation in hematologic diseases.
Humans ; Hematologic Diseases/diagnosis* ; RNA, Messenger/genetics* ; Oncogene Proteins, Fusion/genetics* ; Sequence Analysis, RNA ; Leukemia, Myeloid, Acute/diagnosis*