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.

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.

Objective:To investigate the relationship between the cerebral small vascular disease (CSVD) total burden and the imaging markers and the degree of unilateral middle cerebral artery (MCA) stenosis.Methods:The study was a cross-sectional study. Clinical and imaging data of patients with chronic unilateral MCA stenosis who underwent multimodal MRI from October 2015 to January 2019 in the First Medical Center of PLA General Hospital were retrospectively analyzed. A total of 261 patients were included, 187 males and 74 females. According to the degree of MCA stenosis, the patients were divided into 102 cases in severe stenosis-occlusion group (stenosis degree ≥70%) and 159 cases in mild-moderate stenosis group (stenosis degree <70%). CSVD imaging marker scores (including white matter hyperintensity, perivascular space, cerebral microbleed, and lacune of presumed vascular origin) were assessed according to the ?standards for reporting vascular changes on neuroimaging 1 in the 2 groups, and the CSVD total burden score was calculated. Mann-Whitney U test was used to compare the indicators between the two groups, and the CSVD total burden score and imaging marker scores were ultimately included in a multifactorial binary logistic regression to assess the association of CSVD imaging markers with severe stenosis-occlusion of the MCA after adjusting for vascular risk factors (age, gender, drinking, smoking, hypertension, hyperlipidemia, atrial fibrillation and coronary heart disease). Results:There were significant differences in the CSVD total burden, centrum semiovale perivascular space and lacune of presumed vascular origin score between the mild-to-moderate stenosis group and the severe stenosis-occlusion group (all P<0.05), and none of the differences in the remaining imaging marker scores were statistically significant (all P>0.05). Multivariate binary logistics regression analysis showed CSVD total burden score ( OR=1.300, 95% CI 1.047-1.613, P=0.017), centrum semiovale perivascular space score ( OR=2.099, 95% CI 1.540-2.860, P<0.001) and lacune of presumed vascular origin score ( OR=2.609, 95% CI 1.294-5.261, P=0.007) were independent associated with severe stenosis-occlusion of MCA. Conclusion:The higher CSVD total burden score, centrum semiovale perivascular space score and lacune of presumed vascular origin score are associated with severe stenosis-occlusion of MCA.

Objective:To explore the value of fast susceptibility weighted imaging (SWI) generated by a deep learning model in assessment of acute ischemic stroke (AIS).Methods:From January 2019 to January 2021, 118 AIS patients [75 males and 43 females, aged 23-100 (66±14) years] who underwent MR examination and SWI sequence scanning within 24 h of symptom onset in the First Medical Center of PLA General Hospital were retrospectively analyzed. MATLAB ′s randperm function was used to divide 118 patients into a training set of 96 cases and a test set of 22 cases at a ratio of 8∶2. Fourty-seven AIS patients [38 males and 9 females, aged 16-75 (58±12) years] from one center of a multicenter study were selected to build the external validation set. SWI image and filtered phase image were combined into complex value image as full sampling reference image. Undersampled SWI images were obtained by retrospective undersampling of reference fully sampled images, and the undersampling multiple was five times which could save 80% of the scanning time, then the complex-valued convolutional neural network (ComplexNet) was used to develop reconstruct fast SWI. Interclass correlation coefficient (ICC) or Kappa tests were used to compare the consistency of image quality and the diagnostic consistency for the presence of susceptibility vessel sign (SVS), cerebral microbleeds and asymmetry of cerebral deep medullary veins (DMVs) in AIS patient on fully sampled SWI and fast SWI based on ComplexNet.Results:In test set, score of image quality was 4.5±0.6 for fully sampled SWI image and 4.6±0.7 for fast SWI based on ComplexNet, and coefficient was excellent (ICC=0.86, P<0.05). Full sampling SWI had good agreement with fast SWI based on ComplexNet in detecting SVS (Kappa=0.79, P<0.05), microbleeds (Kappa=0.86, P<0.05), and DMVs asymmetry (Kappa=0.82, P<0.05) in AIS patients. In the external validation set, score of image quality was 4.1±1.0 for fully sampled SWI image and 4.0±0.9 for fast SWI based on ComplexNet, and coefficient was excellent (ICC=0.97, P<0.05). Full sampling SWI had good agreement with fast SWI based on ComplexNet in detecting SVS (Kappa=0.74, P<0.05), microbleeds (Kappa=0.83, P<0.05), and DMVs asymmetry (Kappa=0.74, P<0.05) in AIS patients. Conclusions:Deep learning techniques can significantly accelerate the speed of SWI, and the consistency of image quality and detected AIS signs between fast SWI based on ComplexNet and fully sampled SWI is good. The fast SWI based on ComplexNet can be applied to the radiographic assessment of clinical AIS patients

Objective:To investigate the safety and efficacy of robot-guided ventricular partition puncture drainage in severe intraventricular hemorrhage.Methods:A total of 23 patients with severe intraventricular hemorrhage who underwent robot-guided ventricular partition puncture drainage (experimental group) and 19 patients who underwent robot-guided bilateral ventricular puncture drainage (control group) at Department of Neurosurgery, People's Hospital of Baise from January 2021 to December 2021 were included. The differences in residual hematoma volume within 24 h of surgery, drainage tube retention time, mortality rate within 30 d of surgery, incidence of complications (re-bleeding, intracranial infection, pulmonary infection, hydrocephalus) within 6 months of surgery, and scores of Glasgow coma scale (GCS), activity of daily living (ADL), and National Institutes of Health stroke scale (NIHSS) at 6 months after surgery were compared between the 2 groups.Results:Compared with the control group, the experimental group had significantly lower residual hematoma volume within 24 h of surgery ([8.854±3.519] mL vs. [5.668±2.873] mL), shorter drainage tube retention time ([6.580±1.981] d vs. [4.910±2.763] d), lower incidence of hydrocephalus within 6 months of surgery (42.105% vs. 8.696%), and significantly higher GCS and ADL scores and lower NIHSS scores at 6 months after surgery (8.790±2.898 vs. 11.610±2.948; 69.470±12.899 vs. 78.480±12.861; 13.950±5.265 vs. 9.870±4.124, P<0.05). Conclusion:Robot-guided ventricular partition puncture drainage is a safe and effective surgical method for severe intraventricular hemorrhage.