Zebrafish larvae are useful for identifying chemicals against lateral line (LL) hair cell (HC) damage and this type of chemical screen mainly focuses on searching for protectors against cell death. To expand the candidate pool of HC protectors, a self-built acoustic escape response (AER)-detecting system was developed to apply both low-frequency near-field sound transmission and AER image acquisition/processing modules. The device quickly confirmed the changed LL HC functions caused by most known ototoxins, protectors, and neural transmission modifiers, or knockdown of LL HC-expressing genes. With ten devices wired in tandem, five 'hit' chemicals were identified from 124 cyclin-dependent kinase inhibitors to partially restore cisplatin-damaged AER in less than a day. AS2863619, ribociclib, and SU9516 among the hits, protected the HCs in the mouse cochlea. Therefore, using free-swimming larval zebrafish, the self-made AER-detecting device can efficiently identify compounds that are protective against HC damage, including cell death and loss-of-function.
Animals ; Zebrafish ; Hair Cells, Auditory/physiology* ; Lateral Line System/cytology* ; Escape Reaction/physiology* ; Larva ; Mice ; Cisplatin/toxicity* ; Drug Evaluation, Preclinical/methods*

Objective To explore the value of deep learning(DL)model based on grayscale ultrasound for predicting asymptomatic advanced chronic liver disease(cACLD).Methods Totally 258 patients with asymptomatic compensatory chronic liver diseases were retrospectively included,among them 117 with F3 or F4 stage liver fibrosis were classified into cACLD group,while 141 with F1 or F2 stage liver fibrosis were taken as non-cACLD group.The patients were divided into training set(n=180,including 82 cases of cACLD and 98 cases of non-cACLD)and validation set(n=78,including 35 cases of cACLD and 43 cases of non-cACLD)at the ratio of 7∶3.Univariate and multivariate logistic regression were used to screen independent clinical predictors of cACLD and construct a clinical model.Based on liver grayscale ultrasound,optimal DL features were extracted and screened,and Resnet50 network was adopted as framework,na?ve Bayes classifier was used to construct DL model,and a combined model was constructed based on clinical model and DL model.The efficacy and clinical value of each model for predicting asymptomatic cACLD were evaluated.Results Age,gamma-glutamyl transferase and platelet count were all independent clinical predictors of cACLD,and a clinical model was constructed.Totally 38 optimal DL features were screened to build a DL model.The AUC of combined model in training set and validation set was 0.950 and 0.740,of DL model was 0.944 and 0.737,respectively,being not significantly different(both P＞0.05)but all higher than that of clinical model(0.667 and 0.573,all P＜0.05).Taken 0.59-0.90 as the threshold,the net benefits of combined model in both training and validation sets were higher than that of other models.Conclusion DL model based on grayscale ultrasound could be used to effectively predict asymptomatic cACLD.Combining with clinical characteristics might improve clinical net benefit of this model.

Objective To explore the feasibility and safety of ultrasound-guided percutaneous microwave ablation for unresectable pancreatic cancer.Methods Totally 84 patients who underwent ultrasound-guided percutaneous microwave ablation for unresectable pancreatic cancer were enrolled,and the technical success rate,complete ablation rate,complication rate,pain relief rate and survival time,etc.were observed.Results The median age of 84 cases was 61.5 years.Totally 86 tumors,including 44.19％(38/86)at the head/neck and 55.81％(48/86)at the body/tail of pancreas were detected,and a total of 85 ablation sessions were performed with the median ablation energy applied per tumor of 9.90(1.08,21.60)kJ and the complete ablation rate of 42.86％(36/84).The technical success rate was 100％(85/85).Thirty-nine complication events occurred in 25 cases,no ablation-related death.Among 34 patients underwent ablation mainly for pain symptoms,the pain score decreased from(6.22±1.12)points before treatment to(1.94±1.64)points after treatment(P＜0.001).During 6.8(3.3,12.9)months' follow-up,the mean survival time was(8.5±6.7)months,and all 47 patients died due to tumor progression.Conclusion Ultrasound-guided percutaneous microwave ablation was safe and feasible for unresectable pancreatic cancer.

Purpose To construct a comprehensive model of deep learning features and clinical features based on renal ultrasound for early identification of renal impairment in the pregnant women with preeclampsia.Materials and Methods The information of 279 pregnant women in the First Affiliated Hospital of Guangxi Medical University from January 2018 to June 2023 were retrospectively collected,and all pregnant women were divided the into preeclampsia group(151 cases)and normal group(128 cases).The dataset was randomly divided into a training set(195 samples)and a testing set(84 samples)at a ratio of 7∶3.Based on ultrasound images,the deep learning convolutional neural networks Resnet152 was used to extract deep learning features.The non-zero coefficient features were selected from the deep learning features by the least absolute shrinkage and selection operator,and the K-nearest neighbor algorithm was used to establish the deep learning model.Then,the same classifier model was used to construct a comprehensive model based on clinical data.The receiver operating characteristic curve was used to evaluate the prediction effect.To address the interpretability visualization of models using gradient_weighted class activation mapping and SHapley Additive exPlanations(SHAP)values.Results The area under the curve of the composite model was 0.964(95%CI 0.940-0.988)in the training cohort and 0.899(95%CI 0.835-0.963)in the test cohort.SHAP analysis showed that deep learning features contributed the highest value in the prediction model.Conclusion The comprehensive model based on deep learning combined with clinical features of renal ultrasound can be used to identify renal impairment in normal pregnancy and preeclampsia pregnant women at an early stage,which is conducive to early clinical intervention.

Objective To explore the value of deep learning(DL)model based on grayscale ultrasound for predicting asymptomatic advanced chronic liver disease(cACLD).Methods Totally 258 patients with asymptomatic compensatory chronic liver diseases were retrospectively included,among them 117 with F3 or F4 stage liver fibrosis were classified into cACLD group,while 141 with F1 or F2 stage liver fibrosis were taken as non-cACLD group.The patients were divided into training set(n=180,including 82 cases of cACLD and 98 cases of non-cACLD)and validation set(n=78,including 35 cases of cACLD and 43 cases of non-cACLD)at the ratio of 7∶3.Univariate and multivariate logistic regression were used to screen independent clinical predictors of cACLD and construct a clinical model.Based on liver grayscale ultrasound,optimal DL features were extracted and screened,and Resnet50 network was adopted as framework,na?ve Bayes classifier was used to construct DL model,and a combined model was constructed based on clinical model and DL model.The efficacy and clinical value of each model for predicting asymptomatic cACLD were evaluated.Results Age,gamma-glutamyl transferase and platelet count were all independent clinical predictors of cACLD,and a clinical model was constructed.Totally 38 optimal DL features were screened to build a DL model.The AUC of combined model in training set and validation set was 0.950 and 0.740,of DL model was 0.944 and 0.737,respectively,being not significantly different(both P＞0.05)but all higher than that of clinical model(0.667 and 0.573,all P＜0.05).Taken 0.59-0.90 as the threshold,the net benefits of combined model in both training and validation sets were higher than that of other models.Conclusion DL model based on grayscale ultrasound could be used to effectively predict asymptomatic cACLD.Combining with clinical characteristics might improve clinical net benefit of this model.

Purpose To construct a comprehensive model of deep learning features and clinical features based on renal ultrasound for early identification of renal impairment in the pregnant women with preeclampsia.Materials and Methods The information of 279 pregnant women in the First Affiliated Hospital of Guangxi Medical University from January 2018 to June 2023 were retrospectively collected,and all pregnant women were divided the into preeclampsia group(151 cases)and normal group(128 cases).The dataset was randomly divided into a training set(195 samples)and a testing set(84 samples)at a ratio of 7∶3.Based on ultrasound images,the deep learning convolutional neural networks Resnet152 was used to extract deep learning features.The non-zero coefficient features were selected from the deep learning features by the least absolute shrinkage and selection operator,and the K-nearest neighbor algorithm was used to establish the deep learning model.Then,the same classifier model was used to construct a comprehensive model based on clinical data.The receiver operating characteristic curve was used to evaluate the prediction effect.To address the interpretability visualization of models using gradient_weighted class activation mapping and SHapley Additive exPlanations(SHAP)values.Results The area under the curve of the composite model was 0.964(95%CI 0.940-0.988)in the training cohort and 0.899(95%CI 0.835-0.963)in the test cohort.SHAP analysis showed that deep learning features contributed the highest value in the prediction model.Conclusion The comprehensive model based on deep learning combined with clinical features of renal ultrasound can be used to identify renal impairment in normal pregnancy and preeclampsia pregnant women at an early stage,which is conducive to early clinical intervention.

Objective To explore the feasibility and safety of ultrasound-guided percutaneous microwave ablation for unresectable pancreatic cancer.Methods Totally 84 patients who underwent ultrasound-guided percutaneous microwave ablation for unresectable pancreatic cancer were enrolled,and the technical success rate,complete ablation rate,complication rate,pain relief rate and survival time,etc.were observed.Results The median age of 84 cases was 61.5 years.Totally 86 tumors,including 44.19％(38/86)at the head/neck and 55.81％(48/86)at the body/tail of pancreas were detected,and a total of 85 ablation sessions were performed with the median ablation energy applied per tumor of 9.90(1.08,21.60)kJ and the complete ablation rate of 42.86％(36/84).The technical success rate was 100％(85/85).Thirty-nine complication events occurred in 25 cases,no ablation-related death.Among 34 patients underwent ablation mainly for pain symptoms,the pain score decreased from(6.22±1.12)points before treatment to(1.94±1.64)points after treatment(P＜0.001).During 6.8(3.3,12.9)months' follow-up,the mean survival time was(8.5±6.7)months,and all 47 patients died due to tumor progression.Conclusion Ultrasound-guided percutaneous microwave ablation was safe and feasible for unresectable pancreatic cancer.

Objective To explore the possible role of thrombospondin-1(THBS1)in the excessive fibrosis of vaginal wall induced by estrogen deficiency in rats,the morphological structure of collagen fibers and the expression of THBS1 in the vaginal wall were detected in the estrogen deficiency model of rats.Methods Twenty-four SD rats aged 3 months without delivery were randomly divided into sham operation group and experimental group,with 12 rats in each group.After 12 weeks of modeling,the rats were killed and the vaginal walls were taken for analysis.Masson staining was used to observe the morphological and structural changes of collagen fibers in vaginal wall of rats.Immunohistochemical staining and Western blotting were used to detect the expression level of THBS1 protein.Results After 12 weeks of modeling,the uterine atrophy of experimental group was obvious,the increase of body mass was significantly higher than that of sham operation group,and the level of estradiol was significantly lower than that of sham operation group(P<0.01).Compared with the sham operation group,the upper cortex of vaginal wall of experimental group was significantly atrophy,the smooth muscle bundles were thinner,the muscle gap was wider,the collagen fiber deposition in lamina propria and muscle layer was increased,and the arrangement and distribution were disordered and fragmented.THBS1 expression in vaginal wall of experimental group was significantly higher than that of sham operation group(P<0.05).Conclusion Estrogen deficiency may mediate excessive fibrosis of vaginal wall by upregulating THBS1 expression,thereby damaging the biomechanical properties of vaginal wall and leading to an increased susceptibility to pelvic organ prolapse development.

Objective:To analyze the distribution and drug resistance of wound pathogenic microorganisms in outpatients of wound healing center so as to provide a basis for the standardized construction of wound healing centers.Methods:A retrospective case series study was used to analyzed the data of 365 outpatients treated at Ruijin Hospital, Shanghai Jiaotong University School of Medicine from December 2017 to October 2019. There were 220 males and 145 females, aged (58.8±18.9)years (range, 18-98 years). The patients included 92 first-visit patients and 273 re-visit patients. The culture results (positive rate of pathogenic microorganisms, bacterial species, bacterial distribution) and drug sensitivity results of the wound secretions were compared and analyzed.Results:(1) Among 365 samples of wound secretions, 198 patients were positive for pathogenic microorganisms with a positive rate of 54.3%. A total of 107 strains (51.0%) of Gram-positive bacteria were detected, mainly Staphylococcus aureus (70 strains, 33.3%); 95 strains (45.2%) of Gram-negative bacteria were detected, mainly Escherichia coli (20 strains, 9.5%), followed by Pseudomonas aeruginosa (17 strains, 8.1%); 8 strains (3.8%) of fungi were detected. (2) A total of 26 (28.3%) first-visit patients were positive for pathogenic microorganisms, and 172 (63.0%) re-visit patients were positive for pathogenic microorganisms. The rate of positive microorganism detection had significant differences between first-visit and re-visit patients ( P<0.05). (3) A total of 29 strains were detected in first-visit patients, including 16 strains (55.2%) of Gram-positive bacteria, 11 strains (37.9%) of Gram-negative bacteria and 2 strains (6.9%) of fungi. A total of 181 strains were detected in re-visit patients, including 91 strains (50.3%) of Gram-positive bacteria, 84 strains (46.4%) of Gram-negative bacteria and 6 strains (3.3%) of fungi. The microbial distribution was significantly different between first-visit and re-visit patients ( P<0.05). (4) Compared with first-visit patients, the resistance of Staphylococcus aureus isolated from the re-visit patients to spenicillin, oxacillin, ciprofloxacin, tetracycline, clindamycin, moxifloxacin, erythromycin, and levofloxacin were increased variably. No vancomycin-resistant Staphylococcus aureus was detected, indicating that the staphylococcus aureus presented in the wound was highly sensitive to vancomycin. Conclusions:Staphylococcus aureus is the most common microorganism in wound secretions in outpatients of wound healing center. The rate of positive pathogenic microorganisms in wound secretions of re-visit patients is significantly higher than that of first-visit patients, and the distribution of pathogenic microorganisms of first-visited and revisited patients differs significantly. The Staphylococcus aureus detected in re-visit patients has a higher resistance to common antibiotics compared with first-visit patients. It is suggested that timely detection of pathogenic microorganisms in outpatients and effective control and supervision of outpatient infections are important contents that cannot be ignored in the construction of wound healing center.