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 explore the clinical and imaging characteristics of central nervous system candidiasis due to Candida albicans in immunocompetent infants children.Materials and Methods A retrospective analysis was conducted on immunocompetent children with central nervous system candidiasis diagnosed at Beijing Children's Hospital,Capital Medical University from October 2012 to October 2022,and their clinical and MRI characteristics were analyzed.Results Among the 26 enrolled children,88.5%(23 cases)were under one year old,and 88.5%(23 cases)had high-risk factors.Premature birth was the most common high-risk factor(16 cases,61.5%).All 26 cases had positive MRI examinations,with the main manifestations being meningeal thickening and enhancement(25 cases,96.2%),57.7%(15 cases)involving the skull base meninges,23.1%(six cases)involving the cranial nerves,intracranial localized granulomas(16 cases,61.5%),hydrocephalus(14 cases,53.8%),cerebral infarction(four cases,15.4%),subdural effusion(six cases,23.1%)and cerebral abscess(two cases,7.7%);13 cases of magnetic resonance angiography and magnetic resonance venography examinations were performed,and 10 cases of magnetic resonance angiography abnormalities were found,mainly due to thinning,narrowing or reduced branches of the arterial trunk,and one case of magnetic resonance venography transverse sinus thrombosis.Conclusion Central nervous system candidiasis mostly exists in children younger than one year old.MR imaging shows a more characteristic invasion of the skull base,often causing enhancement of the skull base meninges and cranial nerves,as well as localized granulomas and hydrocephalus.Imaging examination is of great significance for early diagnosis.

Purpose To explore the clinical and imaging characteristics of central nervous system candidiasis due to Candida albicans in immunocompetent infants children.Materials and Methods A retrospective analysis was conducted on immunocompetent children with central nervous system candidiasis diagnosed at Beijing Children's Hospital,Capital Medical University from October 2012 to October 2022,and their clinical and MRI characteristics were analyzed.Results Among the 26 enrolled children,88.5%(23 cases)were under one year old,and 88.5%(23 cases)had high-risk factors.Premature birth was the most common high-risk factor(16 cases,61.5%).All 26 cases had positive MRI examinations,with the main manifestations being meningeal thickening and enhancement(25 cases,96.2%),57.7%(15 cases)involving the skull base meninges,23.1%(six cases)involving the cranial nerves,intracranial localized granulomas(16 cases,61.5%),hydrocephalus(14 cases,53.8%),cerebral infarction(four cases,15.4%),subdural effusion(six cases,23.1%)and cerebral abscess(two cases,7.7%);13 cases of magnetic resonance angiography and magnetic resonance venography examinations were performed,and 10 cases of magnetic resonance angiography abnormalities were found,mainly due to thinning,narrowing or reduced branches of the arterial trunk,and one case of magnetic resonance venography transverse sinus thrombosis.Conclusion Central nervous system candidiasis mostly exists in children younger than one year old.MR imaging shows a more characteristic invasion of the skull base,often causing enhancement of the skull base meninges and cranial nerves,as well as localized granulomas and hydrocephalus.Imaging examination is of great significance for early diagnosis.

OBJECTIVE: To investigate the hepatotoxicity of alkaloids from Euodiae Fructus combined with berberine (BBR) in Zuojin Pill, and to preliminarily explore the possible detoxification mechanism of the combination components. METHODS: The combination ratio of components was determined by the maximum concentration (Cmax) of the chemical components in Zuojin Pill. HepG2 cell model was used to investigate the combined toxicity of the hepatotoxic components from Euodiae Fructus, such as evodiamine (EVO) or dehydroevodiamine (DHED), with BBR for 48 h. The experimental groups were set as follows: the vehicle control group, the EVO group, the DHED group, the BBR group, and the combination group of EVO or DHED with BBR. The cell counting kit-8 (CCK-8) method was used to determine the cell viability, and the combination index (CI) was used to determine the combined toxicity of the components. The alanine transaminase (ALT), aspartate aminotransferase (AST), lactate dehydroge-nase (LDH), and alkaline phosphatase (ALP) activities as well as total bilirubin (TBIL) content in the cell culture supernatant were detected. The protein expression levels of bile acid transporters, such as bile salt export pump (BSEP) and multidrug resistance-associated protein 2 (MRP2), were detected by Western blot. The intracellular malondialdehyde (MDA) content and superoxide dismutase (SOD) activity in HepG2 cells were detected. RESULTS: Compared with EVO or DHED group, the combination of EVO 1 μmol/L with BBR 10 μmol/L or DHED 50 μmol/L with BBR 35 μmol/L significantly increased cell viability of HepG2 cells (P < 0.01), with CI values of 77.89 or 4.49, respectively, much greater than 1. Significant decreases in the activities of ALT, AST, LDH, ALP, and TBIL content in the cell culture supernatant were found in both combination groups (P < 0.05, P < 0.01). Compared with the EVO group, the combination of EVO with BBR upregulated the protein expression levels of BSEP and MRP2. Compared with the DHED group, the combination of DHED with BBR significantly downregulated the protein expression levels of BSEP and MRP2 (P < 0.01). Compared with EVO or DHED group, the combination of EVO or DHED with BBR significantly reduced the MDA content in HepG2 cells (P < 0.05, P < 0.01). CONCLUSION A certain ratio of BBR combined with EVO or DHED had an antagonistic effect on HepG2 cytotoxicity, which might be related to regulating the expression of bile acid transpor-ters, and reducing lipid peroxidation damage.
Humans ; Hep G2 Cells ; Berberine/pharmacology* ; Drugs, Chinese Herbal/toxicity* ; Evodia/chemistry* ; Alkaloids/pharmacology* ; Cell Survival/drug effects* ; Multidrug Resistance-Associated Proteins/metabolism* ; Multidrug Resistance-Associated Protein 2 ; Quinazolines

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.

The activation proteins released by fibroblasts in the tumor microenvironment regulate tumor growth, migration, and treatment response, thereby influencing tumor progression and therapeutic outcomes. Owing to the proliferation and metastasis of tumors, fibroblast activation protein (FAP) is typically highly expressed in the tumor stroma, whereas it is nearly absent in adult normal tissues and benign lesions, making it an attractive target for precision medicine. Radiolabeled agents targeting FAP have the potential for targeted cancer diagnosis and therapy. This comprehensive review aims to describe the evolution of FAPI-based radiopharmaceuticals and their structural optimization. Within its scope, this review summarizes the advances in the use of radiolabeled small molecule inhibitors for tumor imaging and therapy as well as the modification strategies for FAPIs, combined with insights from structure-activity relationships and clinical studies, providing a valuable perspective for radiopharmaceutical clinical development and application.

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 This study aims at establishing a teaching catalog and content for breast rare dis-eases and developing the syllabus for the breast rare disease using integrated multimodality of case-/problem-/resource-based learning(CBL+PBL+RBL).Methods By conducting bibliometrics co-occurrence analysis,we collected 6291 articles on breast rare disease published from January,1975 to June,2024.Additionally,we re-trieved the Textbook on Rare Diseases,the Catalog of Chinese Rare Disease,and Second Batch of Rare Dis-ease Catalog and then decided the teaching content.Results From 16,387 keywords,1000(6.1％)keywords were identified through co-occurrence analysis,including 50(0.3％)candidate diseases.These were classified into three categories:rare primary breast diseases,rare genetic mutation-related diseases associated with breast cancer,and rare systemic multi-system diseases involving the breast.From the candidate list,20(0.1％)rare primary breast diseases were further selected for their notable clinical teaching significance,and significant multi-systemic diseases affecting the breast,whether related to gene mutations or not.Teaching plans were draf-ted using a diversified parallel teaching approaches,taking into account the characteristics of different diseases and the focus of different teaching methods.Conclusions This study initiated the development of the teaching content for breast rare diseases and developed the teaching syllabus using the CBL+PBL+RBL integrated multi teaching model and targeting each rare breast disease for the critical point for teaching.

Purpose To explore the feasibility of targeted diagnosis and localization of prostate cancer via background free differential ultrasound molecular imaging based on prostate-specific membrane antigen (PSMA) targeted ultrasound nanobubbles (NB). Materials and Methods Targeted PSMA-NB and non-targeted NB were constructed. The targeting ability of PSMA-NB on human prostate tumor 22RV1 cells (PSMA positive expression) and PC-3 cells (PSMA negative expression) was determined in vitro. Ten nude mouse models of human prostate tumor 22RV1 cells (n=5) and PC-3 cells (n=5) were constructed. PSMA-NB was injected into the rat tail vein,and in-situ blasting was performed. Ultrasound molecular images before and after blasting were collected,using destruction supplement post-processing technology to obtain and compare the differential ultrasound molecular imaging effects between the two groups. Results The particle size of PSMA-NB and NB were (363.7±24.4) nm and (236.0±55.2) nm,with statistical difference (t=3.19,P=0.007),respectively. Cell targeting results showed that PSMA-NB only adhered to the nucleus with positive PSMA-expression. Animal experiments indicated that the differential ultrasonic molecular images of PSMA positive expression group only showed the highly enhanced area of contrast agent at the tumor site,with no background noise. Conclusion Background free differential ultrasound molecular images can be used for precise targeted diagnosis and localization of PSMA positive prostate cancer,which is constructed based on PSMA targeted ultrasound nanobubbles.