Objective To develop models for predicting response to chemotherapy combined with immunotherapy in patients with esophageal squamous carcinoma with various machine learning algorithms,and then select the optimal model.Methods A retrospective study was performed for 174 patients with esophageal squamous cell carcinoma undergoing chemotherapy combined with immunotherapy admitted in Department of Thoracic Surgery of the First Affiliated Hospital of Army Medical University from January 2022 to December 2023.The CT scans and clinical information were collected before treatment.They were randomly divided into a training set(n=122)and a testing set(n=52)in a ratio of 7∶3.CT radiomic features were extracted and selected,and then 5 machine-learning algorithms were employed to establish the prediction models,including radiomics model and clinical-radiomics model.Five-fold cross-validation was conducted on the training set,and the performance of the prediction models was evaluated on the testing set using receiver operating characteristic(ROC)curve and the F1 score.The best-performing model was further explained using local interpretable model-agnostic explanations(LIME)algorithm.Results Among the 174 patients,115(66.1%)achieved clinical remission.From the clinical information and CT images,1 clinical features and 10 radiomic features were identified.The area under of ROC curve(AUC)for the radiomics and clinical-radiomics models was 0.750(95%CI:0.616～0.883),and 0.766(95%CI:0.637～0.895),respectively.The F1 score of the optimal clinical-radiomics model was 0.829.LIME algorithm indicated that this best model demonstrated reliability in predicting individual samples.Conclusion The clinical-radiomics prediction model based on machine learning algorithm performs well,and can provide a reference for doctors'clinical decision-making by predicting the response to chemotherapy combined with immunotherapy in patients with esophageal squamous cell carcinoma.

There exist risks of ionizing radiation in radiodiagnosis examinations. Implementing shielding protection following the optimization and as low as reasonably achievable (ALARA) principles represents a measure to reduce radiation doses to patients. The implementation of shielding protection in clinical practices should meet high requirements due to variations in the modalities and items in radiodiagnosis examinations, the characteristics and irradiation method of X-ray beams, the method of automatic selection of image quality and radiation dose-related parameters by imaging equipment, the radiation sensitivity of human tissues and organs. This review introduced the shielding products, methods and effects in various radiodiagnosis examinations, as well as the current status and challenges in their applications, aiming to provide a reference for future related research and clinical practices.

Objective:To investigate the diagnostic value of different conventional ultrasound-based radiomics models and their combination with clinical ultrasound features in the pathological subtyping of renal cell carcinoma.Methods:Retrospective data from 286 patients diagnosed with renal cell carcinoma by pathology at the Tenth Affiliated Hospital of Southern Medical University between May 1，2017 and June 7，2024 were collected. Among the 286 patients，203 were clear cell carcinoma，44 were papillary renal cell carcinoma，and 39 were chromophobe renal cell carcinoma. The patients were randomly divided into a training group（201 cases）and a validation group（85 cases）in a ratio of 7 to 3. Regions of interest（ROI）were delineated on conventional ultrasound images，and the radiomics features were extracted. Feature selection was performed using Student's t-test，Pearson correlation，and the least absolute shrinkage and selection operator（LASSO）. Six different machine learning methods included category gradient boosting（CatBoost），light gradient boosting machine（LightGBM），Logistic regression（LR），random forest（RF），support vector machine（SVM）and extreme gradient boosting（XGBoost）were used to establish radiomics models. Weight balancing was applied to correct for sample imbalance，and an imaging genomics model was constructed after balancing the samples. Independent predictors of renal cell carcinoma subtyping were selected from clinical ultrasound features using univariate and multivariate logistic regression analyses，and a clinical imaging model was constructed. The best-performing radiomics model was combined with the clinical independent predictors to construct a combined model. Receiver operating characteristic curves and the obuchowski index were plotted to evaluate model performance. Results:Among the radiomics models，the model constructed using Random Forest（RS RF）after balancing the samples exhibited the best predictive performance，with area under the curve（AUCs）of 0.918（micro-average ROC）and 0.903（macro-average ROC），and the obuchowski index was 0.885 in the validation group. The long and short axes of ultrasound image tumor masses were used as imaging independent predictors to construct a clinical imaging model. In the validation group，the AUCs of the clinical model were 0.886（micro-average ROC）and 0.606（macro-average ROC），and the obuchowski index was 0.569. The combined model achieved AUCs of 0.888（micro-average ROC）and 0.967（macro-average ROC），with an obuchowski index of 0.933，outperforming any single model. Conclusions:The combination of conventional ultrasound-based radiomics models with clinical ultrasound features demonstrates high diagnostic value in differentiating clear cell carcinoma，papillary renal cell carcinoma，and chromophobe renal cell carcinoma. It may serve as an auxiliary tool for providing timely and effective clinical guidance.

Large language models (LLMs) such as ChatGPT, Claude, Llama, and Qwen are emerging as transformative technologies for the diagnosis and treatment of various diseases. With their exceptional long-context reasoning capabilities, LLMs are proficient in clinically relevant tasks, particularly in medical text analysis and interactive dialogue. They can enhance diagnostic accuracy by processing vast amounts of patient data and medical literature and have demonstrated their utility in diagnosing common diseases and facilitating the identification of rare diseases by recognizing subtle patterns in symptoms and test results. Building on their image-recognition abilities, multimodal LLMs (MLLMs) show promising potential for diagnosis based on radiography, chest computed tomography (CT), electrocardiography (ECG), and common pathological images. These models can also assist in treatment planning by suggesting evidence-based interventions and improving clinical decision support systems through integrated analysis of patient records. Despite these promising developments, significant challenges persist regarding the use of LLMs in medicine, including concerns regarding algorithmic bias, the potential for hallucinations, and the need for rigorous clinical validation. Ethical considerations also underscore the importance of maintaining the function of supervision in clinical practice. This paper highlights the rapid advancements in research on the diagnostic and therapeutic applications of LLMs across different medical disciplines and emphasizes the importance of policymaking, ethical supervision, and multidisciplinary collaboration in promoting more effective and safer clinical applications of LLMs. Future directions include the integration of proprietary clinical knowledge, the investigation of open-source and customized models, and the evaluation of real-time effects in clinical diagnosis and treatment practices.
Humans ; Large Language Models ; Tomography, X-Ray Computed

Objective:To investigate the diagnostic value of different conventional ultrasound-based radiomics models and their combination with clinical ultrasound features in the pathological subtyping of renal cell carcinoma.Methods:Retrospective data from 286 patients diagnosed with renal cell carcinoma by pathology at the Tenth Affiliated Hospital of Southern Medical University between May 1，2017 and June 7，2024 were collected. Among the 286 patients，203 were clear cell carcinoma，44 were papillary renal cell carcinoma，and 39 were chromophobe renal cell carcinoma. The patients were randomly divided into a training group（201 cases）and a validation group（85 cases）in a ratio of 7 to 3. Regions of interest（ROI）were delineated on conventional ultrasound images，and the radiomics features were extracted. Feature selection was performed using Student's t-test，Pearson correlation，and the least absolute shrinkage and selection operator（LASSO）. Six different machine learning methods included category gradient boosting（CatBoost），light gradient boosting machine（LightGBM），Logistic regression（LR），random forest（RF），support vector machine（SVM）and extreme gradient boosting（XGBoost）were used to establish radiomics models. Weight balancing was applied to correct for sample imbalance，and an imaging genomics model was constructed after balancing the samples. Independent predictors of renal cell carcinoma subtyping were selected from clinical ultrasound features using univariate and multivariate logistic regression analyses，and a clinical imaging model was constructed. The best-performing radiomics model was combined with the clinical independent predictors to construct a combined model. Receiver operating characteristic curves and the obuchowski index were plotted to evaluate model performance. Results:Among the radiomics models，the model constructed using Random Forest（RS RF）after balancing the samples exhibited the best predictive performance，with area under the curve（AUCs）of 0.918（micro-average ROC）and 0.903（macro-average ROC），and the obuchowski index was 0.885 in the validation group. The long and short axes of ultrasound image tumor masses were used as imaging independent predictors to construct a clinical imaging model. In the validation group，the AUCs of the clinical model were 0.886（micro-average ROC）and 0.606（macro-average ROC），and the obuchowski index was 0.569. The combined model achieved AUCs of 0.888（micro-average ROC）and 0.967（macro-average ROC），with an obuchowski index of 0.933，outperforming any single model. Conclusions:The combination of conventional ultrasound-based radiomics models with clinical ultrasound features demonstrates high diagnostic value in differentiating clear cell carcinoma，papillary renal cell carcinoma，and chromophobe renal cell carcinoma. It may serve as an auxiliary tool for providing timely and effective clinical guidance.

There exist risks of ionizing radiation in radiodiagnosis examinations. Implementing shielding protection following the optimization and as low as reasonably achievable (ALARA) principles represents a measure to reduce radiation doses to patients. The implementation of shielding protection in clinical practices should meet high requirements due to variations in the modalities and items in radiodiagnosis examinations, the characteristics and irradiation method of X-ray beams, the method of automatic selection of image quality and radiation dose-related parameters by imaging equipment, the radiation sensitivity of human tissues and organs. This review introduced the shielding products, methods and effects in various radiodiagnosis examinations, as well as the current status and challenges in their applications, aiming to provide a reference for future related research and clinical practices.

Objective To understand the change trend and characteristics of nosocomial infection through the comprehensive surveillance on nosocomial infection in infectious diseases specialized hospitals dur-ing 2017-2023 to provide an evidence for the prevention,control and management of nosocomial infection.Methods The surveillance indicators of nosocomial infection in a hospital during 2017-2023 were collected.The nosocomial infection rate,nosocomial infection rate in different inpatient wards,nosocomial infection sites,nosocomial infection pathogenic bacterial distribution and susceptibility factors conducted the statistical analysis.Results A total of 93 254 patients were admitted and treated during 2017-2023.The nosocomial in-fection rate and infection case-times rate showed the decreasing trend(P＜0.05).The case-times rate of the patients in the AIDS wards was 3.75％,which was higher than 0.79％in the tuberculosis wards(P＜0.05).The nosocomial infection case-times rate in the two wards areas during 2017-2023 showed the decreasing trend(P＜0.05).The infection sites were mainly the respiratory system(61.56％),blood system(9.44％)and urinary system(8.61％).A total of 803 strains of pathogenic bacteria were detected out,which were mainly Gram negative bacteria(63.89％),the top five were in turn Klebsiella pneumoniae(16.19％),Esche-richia coli(15.57％),Acinetobacter baumannii(10.83％),Pseudomonas aeruginosa(8.84％)and Staphylo-coccus aureus(7.22％).The top three of susceptibility factors were low immune function(58.78％),long term antibiotic use(11.29％)and ventilator use(9.20％).Conclusion Initiatively carrying the hospital infec-tion surveillance could accurately grasp the incidence trends and provide the direction and data support for pre-vention and control priorities.

There is a shared problem in current optical imaging technologies of how to obtain the optical parameters of biological tissues with complex profiles. In this work, an imaging system for obtaining the optical parameters of biological tissues with complex profile was presented. Firstly, Fourier transformation profilometry was used for obtaining the profile information of biological tissues, and then the difference of incident light intensity at different positions on biological tissue surface was corrected with the laws of illumination, and lastly the optical parameters of biological tissues were achieved with the spatial frequency domain imaging technique. Experimental results indicated the proposed imaging system could obtain the profile information and the optical parameters of biological tissues accurately and quickly. For the slab phantoms with height variation less than 30 mm and angle variation less than 40º, the maximum relative errors of the profile uncorrected optical parameters were 46.27% and 72.18%, while the maximum relative errors of the profile corrected optical parameters were 6.89% and 10.26%. Imaging experiments of a face-like phantom and a human's prefrontal lobe were performed respectively, which demonstrated the proposed imaging system possesses clinical application value for the achievement of the optical parameters of biological tissues with complex profiles. Besides, the proposed profile corrected method can be used to combine with the current optical imaging technologies to reduce the influence of the profile information of biological tissues on imaging quality.
Diagnostic Imaging ; Humans ; Light ; Optical Imaging ; Phantoms, Imaging

Cow′s milk protein allergy (CMPA) is one of the most common presentations of food allergy seen in early childhood.It is an abnormal immune response caused by cow′s milk protein.CMPA can be clinically subdivided into either immediate-onset IgE mediated or delayed onset non-IgE mediated, or both.At present, concerns regarding the early and timely diagnosis of CMPA have been high-lighted over the years and there are many expert consensus on CMPA in China, but these consensus did not distinguish IgE mediated or non-IgE mediated CMPA.In view of the obvious clinical differences between the two type of CMPA and non-IgE mediated CMPA is more common in infancy, experts focus on pediatric gastroenterology, allergy/immunology, dermatology, nutrition and child healthcare convened by the Allergy Prevention and Control Professional Committee of Chinese Preventive Medicine Association present this guideline to help practitioners in primary care settings to early recognize and make suitable management of non-IgE mediated CMPA in China.The guideline incorporates the cutting-edge international guidance and the actual situation of Chinese children describing in detail the types, clinical features, diagnosis and nutritional intervention of non-IgE mediated CMPA.There are 42 recommendations in 7 categories in total referring to the common questions related to non-IgE mediated CMPA.

Objective To study the mutation of STK11 gene in a Chinese family and a sporadic patient with Peutz-Jeghers syndrome (PJS),and to provide a basis for genetic diagnosis and counseling.Methods One sporadic patient and two patients from a family with PJS were collected,all of whom had typical mucosal pigmentation and gastrointestinal polyposis.Blood samples were obtained from the two patients and six unaffected relatives in this family,the sporadic patient,and 100 healthy controls.DNA was extracted,and PCR was performed to amplify nine exons and their adjacent introns in the STK11 gene followed by direct sequencing.The sequencing results were aligned to the published sequence of STK11 gene from Genbank.Results No mutation was found in the STK11 gene of any of the patients,unaffected relatives,or healthy controls.Conclusions Genetic heterogeneity exists in Peutz-Jeghers syndrome,hinting that there may be other causative genes or sites for this entity.