Ultrasound medicine is crucial to assist clinical diagnosis and treatment. The application of artificial intelligence in ultrasound medicine has received extensive attention to assist in clinical diagnosis and improve diagnostic accuracy and prognosis. However，the generalization of existing models is limited by small sample size，data heterogeneity，and patient privacy protection. Federated learning，as a distributed learning paradigm，enables multiple centers to conduct local training and aggregate model parameters to jointly train a global model，effectively increasing the sample size and data diversity without exchanging raw data，thereby protecting patient privacy. This approach has promising clinical application prospects. However，there are still challenges in optimizing the defense capability，performance，and diverse applicability of the model. This article reviews the application and challenges of federated learning in ultrasound image analysis and diseases diagnosis.

Acute symptomatic osteoporotic thoracolumbar compression fracture (ASOTLF) can lead to chronic low back pain, kyphosis deformity, pulmonary dysfunction, loss of mobility, and even life-threatening complications. Vertebral augmentation is currently the mainstream treatment method for this condition. In 2019, the Editorial Board of Chinese Journal of Trauma and the Spinal Trauma Group of Orthopedic Surgeons Branch of Chinese Medical Doctor Association collaboratively led the development of Clinical guideline for vertebral augmentation for acute symptomatic osteoporotic thoracolumbar compression fractures. Six years later, with advances in clinical diagnosis and treatment techniques as well as accumulating evidence in related fields, the 2019 guideline requires updating. To this end, the Spinal Trauma Group of Orthopedic Surgeons Branch of Chinese Medical Doctor Association, the Spinal Health Professional Committee of China Human Health Science and Technology Promotion Association, and the Minimally Invasive Orthopedics Professional Committee of Shaanxi Medical Doctor Association have organized experts in the field to develop the Clinical guideline for vertebral augmentation of acute symptomatic osteoporotic thoracolumbar compression fractures ( version 2025) , based on the latest evidence-based medical researches. This guideline incorporates 3 recommendations retained from the 2019 version with updated strength of evidence, along with 12 new recommendations. It provides recommendations from six aspects of diagnosis, pain management, treatment option selection, prevention of postoperative complications, anti-osteoporosis therapy, and postoperative rehabilitation, aiming to provide a reference for standard treatment of vertebral augmentation for ASOTLF in hospitals at all levels.

Objective To investigate the effects of Tongyangxiao(TYX)lotion on the healing of infected wounds and the expression of interleukin(IL)-1β and inhibitor of nuclear factor κBα(IκBα)/p65 in rats.Methods Fifty rats were randomly assigned to the model group,potassium permanganate(PP)group,and low-,medium-,and high-dose(TYX-L,TYX-M,TYX-H)groups,with 10 rats in each group.An open infection model of full-thickness skin defects was established,and the rats in each group were treated with the corresponding medicinal solution for dressing changes once a day for a total of 14 days.The wound healing of rats was observed,and immunofluorescence,immunohistochemistry,and Western blotting were used to detect the nuclear translocation rate of p65 and the expression levels of IL-1β,p-IκBα/IκBα,and p-p65/p65 proteins in the granulation tissue.Results Compared with the model group,the wound healing rates of both the TYX-M group and the TYX-H group increased on the 7th day of treatment(P<0.05).Compared with the model group and the PP group,the wound healing rates of the TYX-L group,the TYX-M group and the TYX-H group increased on the 14th day of treatment(P<0.05).On the 7th day of treatment,the expression of IL-1β in the TYX-M group was lower than that in the PP group and the Model group(P<0.05),and the p-IκBα/IκBα ratio in the TYX-H group was lower than that in the PP group and the model group(P<0.05).The nuclear translocation rates of p65 in the TYX-L group,TYX-M group and TYX-H group were lower than those in the PP group and model group(P<0.05).On the 14th day of treatment,the p-p65/p65 ratio in the TYX-H group and the TYX-M group was lower than that in the model group(P<0.05),and the IL-1β in the TYX-L group,the TYX-M group,and the TYX-H group and p-IκBα/IκBα ratio in the TYX-H group were lower than those in the PP group and the model group(P<0.05).Conclusion The mechanism of TYX lotion in promoting the healing of infected wounds was associated with the suppression of the activation of NF-κB signaling pathway and alleviation of inflammatory responses.

Metagenomic next-generation sequencing, as an unbiased detection technology, demonstrates higher diagnostic efficacy than traditional methods. Gut microorganisms are important flora for safeguarding health and have become a hot research topic. Modeling and analyzing the genomic data of intestinal flora using machine learning is very important in disease prediction and diagnosis. This paper briefly introduces the characteristics of metagenomic next-generation sequencing, key algorithms and evaluation indexes of machine learning, outlines the main steps of combining machine learning with metagenomic next-generation sequencing, and summarizes the application of the combination of machine learning and metagenomic next-generation sequencing technology in the study of intestinal flora, which will provide a more accurate method for diagnosis and prediction of the related diseases, and give more ideas for the future research and clinical practice.

With the rapid development of computer vision technology, deep learning models have demonstrated new research area and potential value in intelligent microbiological detection. By utilizing multilayer neural networks and large amounts of training data, these models are capable of automated extraction and analysis for complex features, thereby improving the efficiency and accuracy of detection. This paper introduces the research background of deep learning in microbiological image detection, and elaborates on the methods for constructing microbiological image datasets, including data collection, preprocessing, annotation, and partitioning, and introduces typical deep learning models as well as their application examples in various microbiological detection. Deep learning in microbiological image analysis faces numerous challenges which needs further development.

Objective:This study aims to develop a rapid identification technique for various genotypes of extended-spectrum β-lactamase (ESBL) producing Escherichia coli using matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) in conjunction with machine learning algorithms. Methods:A total of 158 Escherichia coli strains were isolated from the clinical laboratory of the First Hospital of Jilin University from August 2018 to December 2022. Polymerase chain reaction (PCR) was employed to detect the CTX-M-1, CTX-M-8, CTX-M-9, and SHV genes. Mass spectral data of the bacterial strains were acquired by MALDI-TOF MS with a cooperative matrix of (E)-propyl α-cyano-4-hydroxycinnamate (CHCA-C3). Models based on random forest (RF), logistic regression (LR), and support vector machine (SVM) algorithms were constructed. The performance of the constructed models was evaluated using metrics including accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC). Mass spectral peaks exhibiting sensitivity and specificity exceeding 80% in the models were designated as characteristic peaks. To validate the efficacy of the cooperative matrix of CHCA-C3, clinical isolates of ESBL-producing Escherichia coli were analyzed by MALDI-TOF MS using the conventional CHCA matrix for comparative purposes. Results:Among the 158 strains of Escherichia coli, 91 strains produced ESBL, all of which were CTX-M genotype. The AUC values for the respective models were as follows: CTX-M-1 genotype exhibited AUC values of 0.98 for LR, 1.00 for RF, and 0.73 for SVM; CTX-M-9 genotype exhibited AUC values of 0.93 for LR, 0.99 for RF, and 0.76 for SVM; for CTX-M-8, all models achieved an AUC of 1.00, indicating excellent classification performance with respect to accuracy, specificity, and sensitivity. The characteristic mass spectral peaks associated with each genotype included: CTX-M-1 genotype at m/z 6 390; CTX-M-8 genotype at m/z 5 224, m/z 5 393, and m/z 9 021; CTX-M-9 genotype at m/z 5 161 and m/z 5 273. In the MALDI-TOF MS analysis conducted with the conventional CHCA matrix, the characteristic peak at m/z 9 021 for CTX-M-8 was the only one detected, with the characteristic peaks for CTX-M-1 and CTX-M-9 remaining undetected. Conclusion:The application of cooperative matrix of CHCA-C3 in conjunction with MALDI-TOF MS and machine learning algorithms facilitates the rapid and precise identification of extended-spectrum β-lactamase (ESBL)-producing Escherichia coli. This approach offers a feasible solution for evidence-based clinical therapy and the control of healthcare-associated infections.

Acute symptomatic osteoporotic thoracolumbar compression fracture (ASOTLF) can lead to chronic low back pain, kyphosis deformity, pulmonary dysfunction, loss of mobility, and even life-threatening complications. Vertebral augmentation is currently the mainstream treatment method for this condition. In 2019, the Editorial Board of Chinese Journal of Trauma and the Spinal Trauma Group of Orthopedic Surgeons Branch of Chinese Medical Doctor Association collaboratively led the development of Clinical guideline for vertebral augmentation for acute symptomatic osteoporotic thoracolumbar compression fractures. Six years later, with advances in clinical diagnosis and treatment techniques as well as accumulating evidence in related fields, the 2019 guideline requires updating. To this end, the Spinal Trauma Group of Orthopedic Surgeons Branch of Chinese Medical Doctor Association, the Spinal Health Professional Committee of China Human Health Science and Technology Promotion Association, and the Minimally Invasive Orthopedics Professional Committee of Shaanxi Medical Doctor Association have organized experts in the field to develop the Clinical guideline for vertebral augmentation of acute symptomatic osteoporotic thoracolumbar compression fractures ( version 2025) , based on the latest evidence-based medical researches. This guideline incorporates 3 recommendations retained from the 2019 version with updated strength of evidence, along with 12 new recommendations. It provides recommendations from six aspects of diagnosis, pain management, treatment option selection, prevention of postoperative complications, anti-osteoporosis therapy, and postoperative rehabilitation, aiming to provide a reference for standard treatment of vertebral augmentation for ASOTLF in hospitals at all levels.

Metagenomic next-generation sequencing, as an unbiased detection technology, demonstrates higher diagnostic efficacy than traditional methods. Gut microorganisms are important flora for safeguarding health and have become a hot research topic. Modeling and analyzing the genomic data of intestinal flora using machine learning is very important in disease prediction and diagnosis. This paper briefly introduces the characteristics of metagenomic next-generation sequencing, key algorithms and evaluation indexes of machine learning, outlines the main steps of combining machine learning with metagenomic next-generation sequencing, and summarizes the application of the combination of machine learning and metagenomic next-generation sequencing technology in the study of intestinal flora, which will provide a more accurate method for diagnosis and prediction of the related diseases, and give more ideas for the future research and clinical practice.

With the rapid development of computer vision technology, deep learning models have demonstrated new research area and potential value in intelligent microbiological detection. By utilizing multilayer neural networks and large amounts of training data, these models are capable of automated extraction and analysis for complex features, thereby improving the efficiency and accuracy of detection. This paper introduces the research background of deep learning in microbiological image detection, and elaborates on the methods for constructing microbiological image datasets, including data collection, preprocessing, annotation, and partitioning, and introduces typical deep learning models as well as their application examples in various microbiological detection. Deep learning in microbiological image analysis faces numerous challenges which needs further development.

Objective:This study aims to develop a rapid identification technique for various genotypes of extended-spectrum β-lactamase (ESBL) producing Escherichia coli using matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) in conjunction with machine learning algorithms. Methods:A total of 158 Escherichia coli strains were isolated from the clinical laboratory of the First Hospital of Jilin University from August 2018 to December 2022. Polymerase chain reaction (PCR) was employed to detect the CTX-M-1, CTX-M-8, CTX-M-9, and SHV genes. Mass spectral data of the bacterial strains were acquired by MALDI-TOF MS with a cooperative matrix of (E)-propyl α-cyano-4-hydroxycinnamate (CHCA-C3). Models based on random forest (RF), logistic regression (LR), and support vector machine (SVM) algorithms were constructed. The performance of the constructed models was evaluated using metrics including accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC). Mass spectral peaks exhibiting sensitivity and specificity exceeding 80% in the models were designated as characteristic peaks. To validate the efficacy of the cooperative matrix of CHCA-C3, clinical isolates of ESBL-producing Escherichia coli were analyzed by MALDI-TOF MS using the conventional CHCA matrix for comparative purposes. Results:Among the 158 strains of Escherichia coli, 91 strains produced ESBL, all of which were CTX-M genotype. The AUC values for the respective models were as follows: CTX-M-1 genotype exhibited AUC values of 0.98 for LR, 1.00 for RF, and 0.73 for SVM; CTX-M-9 genotype exhibited AUC values of 0.93 for LR, 0.99 for RF, and 0.76 for SVM; for CTX-M-8, all models achieved an AUC of 1.00, indicating excellent classification performance with respect to accuracy, specificity, and sensitivity. The characteristic mass spectral peaks associated with each genotype included: CTX-M-1 genotype at m/z 6 390; CTX-M-8 genotype at m/z 5 224, m/z 5 393, and m/z 9 021; CTX-M-9 genotype at m/z 5 161 and m/z 5 273. In the MALDI-TOF MS analysis conducted with the conventional CHCA matrix, the characteristic peak at m/z 9 021 for CTX-M-8 was the only one detected, with the characteristic peaks for CTX-M-1 and CTX-M-9 remaining undetected. Conclusion:The application of cooperative matrix of CHCA-C3 in conjunction with MALDI-TOF MS and machine learning algorithms facilitates the rapid and precise identification of extended-spectrum β-lactamase (ESBL)-producing Escherichia coli. This approach offers a feasible solution for evidence-based clinical therapy and the control of healthcare-associated infections.