Artificial intelligence methods are developing rapidly in the medical field.However,the effectiveness of model training relies heavily on the support of sufficient sample sizes.Due to various constraints such as privacy,security,ethics,and costs in the medical field,it is rather difficult to obtain a large number of labeled training samples.Problems like the scarcity of rare disease cases,the lack of biological data for drug molecule mining,and the shortage of high-quality annotations for medical images significantly reduce the ability of models to learn from observed data,which in turn leads to poor prediction performance.In this context,constructing efficient learning artificial intelligence models for small sample data is of far-reaching significance both theoretically and practically.On the one hand,it can help to explore potential patterns when samples are insufficient in the early stage of new research.On the other hand,high-quality models can effectively reduce the cost of manual annotation,shorten the research cycle,and provide opportunities for solving challenging problems in medical research where it is difficult to collect a sufficient number of samples.Driven by both the expected advantages and actual needs,the research on artificial intelligence for small sample data has gradually become a highly anticipated and important research direction.This review systematically collates and summarizes the principles,advantages,disadvantages,applicable scenarios,and principal challenges associated with six artificial intelligence methods currently employed in the context of small-sample medical data,namely generative adversarial networks,graph neural networks,transfer learning,reinforcement learning,and Meta-learning.Furthermore,the review provides an extensive outlook and in-depth contemplation on the future trajectory of artificial intelligence methodologies in the realm of small sample data in medicine.

Artificial intelligence methods are developing rapidly in the medical field.However,the effectiveness of model training relies heavily on the support of sufficient sample sizes.Due to various constraints such as privacy,security,ethics,and costs in the medical field,it is rather difficult to obtain a large number of labeled training samples.Problems like the scarcity of rare disease cases,the lack of biological data for drug molecule mining,and the shortage of high-quality annotations for medical images significantly reduce the ability of models to learn from observed data,which in turn leads to poor prediction performance.In this context,constructing efficient learning artificial intelligence models for small sample data is of far-reaching significance both theoretically and practically.On the one hand,it can help to explore potential patterns when samples are insufficient in the early stage of new research.On the other hand,high-quality models can effectively reduce the cost of manual annotation,shorten the research cycle,and provide opportunities for solving challenging problems in medical research where it is difficult to collect a sufficient number of samples.Driven by both the expected advantages and actual needs,the research on artificial intelligence for small sample data has gradually become a highly anticipated and important research direction.This review systematically collates and summarizes the principles,advantages,disadvantages,applicable scenarios,and principal challenges associated with six artificial intelligence methods currently employed in the context of small-sample medical data,namely generative adversarial networks,graph neural networks,transfer learning,reinforcement learning,and Meta-learning.Furthermore,the review provides an extensive outlook and in-depth contemplation on the future trajectory of artificial intelligence methodologies in the realm of small sample data in medicine.

Objective To investigate the predictive value of integrating clinical pathological characteristics with genetic testing for occult lymph node metastasis(OLNM)in patients with papillary thyroid microcarcinoma(PTMC).Methods A total of 104 PTMC patients admitted to our hospital between May 2023 and May 2025 were included in the study.All patients showed no evidence of suspicious lymph node metastasis on preoperative imaging and underwent standard thyroidectomy with central lymph node dissection.Based on postoperative pathological con-firmation of central lymph node metastasis status,patients were classified into an OLNM-positive group(n=53)and an OLNM-negative group(n=51).Baseline characteristics,clinicopathological features,BRAFV600E gene mutation status,and TERT promoter mutation status were compared between the two groups.To identify factors independently associated with OLNM in PTMC patients,multivariate logistic regression analysis was conducted.The area under the receiver operating characteristic curve(AUC)was utilized to assess the predictive performance of a combined model incorporating clinical,pathological,and genetic features for OLNM.Results Compared with the OLNM-negative group,the OLNM-positive group exhibited significantly higher preoperative thyroid-stimulating hormone(TSH)levels(P<0.05).Moreover,the OLNM-positive group demonstrated significantly greater proportions of tumors with diameter>0.5 cm,multifocality,microcalcifications,capsule invasion,extrathyroidal extension,T3 stage,BRAFV600E mutation,and TERT promoter mutation(all P<0.05).Multivariate logistic regression analysis identified preoperative TSH level,tumor diameter>0.5 cm,multifocal lesions,capsule invasion,extrathyroidal extension,T stage,BRAFV600E mutation,and TERT promoter mutation as independent risk factors for OLNM in patients with PTMC(all P<0.05).ROC curve analysis demonstrated that the integrated model combining clinical pathological features-including tumor diameter,number of lesions,microcalcification,capsule invasion,extrathy-roidal extension,and T stage-with genetic markers(BRAFV600E and TERT promoter mutations)exhibited the highest predictive performance,yielding an AUC of 0.940.This was significantly higher than the model based solely on clinical pathological features(AUC=0.736)or those relying exclusively on genetic testing(BRAFV600E:AUC=0.860;TERT:AUC=0.882),with all comparisons reaching statistical significance(P<0.05).Conclusions The integration of clinical pathological features with genetic testing significantly improved the predictive accuracy of OLNM in PTMC patients,surpassing models based solely on individual clinical pathological characteristics or genetic tests alone.This multimodal strategy offers a robust,evidence-based foundation for personalized surgical planning and enhances the precision of clinical decision-making in the management of PTMC.

Objective:To explore the clinical distribution characteristics and changes of common inhalant allergens and food allergens in all outpatient and inpatient children visiting Shanghai Children′s Hospital from 2020 to 2024, and to provide a basis for the diagnosis, treatment and prevention of allergic diseases in children.Methods:A retrospective cohort study was conducted to retrospectively enroll all outpatient and inpatient children who visited Shanghai Children′s Hospital and underwent serum allergen-specific IgE (sIgE) antibody testing from January 2020 to August 2024, and the characteristics and changing trends of allergens in the past 5 years were analyzed. A total of 127 310 tests were included. There were 76 776 male tests (60.31%) and 50 534 female tests (39.69%). There were 27 392 tests (21.52%) aged 0-3 years (infant group), 51 596 tests (40.53%) aged 4-6 years (preschool group), 44 574 tests (35.01%) aged 7-12 years (school-age group), and 3 748 tests (2.94%) aged 13-18 years (adolescent group). The χ2 test was used for statistical analysis. Results:The difference in total positivity rate between different years was statistically significant ( χ2=2 907.478, P<0.001). The positive rates of inhalant allergens such as house dust, Dermatophagoides pteronyssinus/ Dermatophagoides farinae, cockroach, cat dander, mugwort, Humulus scandens, mold fungi mix, and food allergens such as beef and mutton increased significantly with age (The χ2 values were 649.496, 3 414.686, 303.247, 1 277.408, 40.477, 189.952, 600.737, 203.198, and 15.301, respectively, and the P values were <0.001, <0.001, <0.001, <0.001, <0.001, <0.001,<0.001,<0.001, and 0.002, respectively). The positive rates of inhalant allergen such as Ambrosia elatior (1.59%) and food allergens such as egg white (17.36%), milk (30.48%), shrimp (8.27%), crab (8.13%), codfish (2.61%), salmon (0.66%), mytilus edulis (2.89%), lobster/scallop (5.27%), cashew nuts (5.09%), peanuts (3.54%), and soybean (1.73%) were highest at the age of 0-3 years and decreased significantly with age (The χ2 values were 10.365, 2 407.443, 139.085, 872.548, 870.245, 106.823, 47.674, 47.244, 559.422, 369.800, 384.788, 153.660, respectively, and the P values were 0.016, 0.000,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001, respectively). Inhaled allergens mainly have a greater impact on children with respiratory-related diseases such as allergic rhinitis and asthma, while food allergens mainly have a greater impact on children with atopic dermatitis/eczema. The positive rate of sIgE of various allergens in the allergic rhinitis combined asthma group were higher than that of allergic rhinitis alone, and the sIgE positive rate of total allergens and inhaled allergens was significantly higher than that of allergic rhinitis alone ( χ2=20.851, 39.155, the P values were both<0.001). Among them, the sIgE positive rate of Ambrosia elatior and cashew nuts showed significant difference ( χ2=5.044, 8.420, P=0.025, 0.004); and the sIgE positive rate of Dermatophagoides pteronyssinus/ Dermatophagoides farinae, cat dander, grass pollens mix and mold fungi mix had extremely significant difference ( χ2=26.409, 25.990, 21.283, 16.411, the P values were all <0.001). The inhaled allergens and food allergens with the highest positive rates in the 5 years were Dermatophagoides pteronyssinus/ Dermatophagoides farinae (56.21%) and milk (47.47%), and as time went by, the positive rates gradually decreased. There is a moderate correlation between the three allergens of Ambrosia elatior, Amaranthus retroflexus, and tree pollens mix (0.55, 0.70, 0.63), and there is a moderate correlation between mango and tree pollens mix (0.50). Conclusion:Dermatophagoides pteronyssinus/ Dermatophagoides farinae, cat dander, dog dander, egg white, and milk may be important allergens for children in Shanghai City from 2020 to 2024. The positive rates vary among different genders, age groups, and disease groups, but the positive rates of Dermatophagoides pteronyssinus/ Dermatophagoides farinae, milk and cat dander allergens remain in the top three.

Gongfa is an essential approach to prevent and treat diseases in traditional Chinese medicine(TCM),often used to prevent a disease before it arises.Guided by TCM and modern scientific theories,the Tuina(Chinese therapeutic massage)Gongfa prescription theory implements the principle,method,prescription,and form in clinical pattern-identified treatment to prescribe the corresponding Tuina Gongfa prescription,i.e.,to prescribe a basic Tuina Gongfa prescription,specifically for a systemic disease,and modify Gongfa forms based on the basic prescription according to different patterns.The Gongfa prescription for pulmonary diseases designs corresponding Gongfa forms from six perspectives:lifting Yang,securing the exterior,opening the orifices,soothing the chest,harmonizing the stomach,and regulating Qi to prevent and treat diseases.The application of the pulmonary Gongfa prescription indicates the potential to apply the Tuina Gongfa prescription theory for the clinical prevention,treatment,and rehabilitation of disorders of other systems,thereby fully realizing the unique role of TCM Gongfa.

Objective To investigate the predictive value of integrating clinical pathological characteristics with genetic testing for occult lymph node metastasis(OLNM)in patients with papillary thyroid microcarcinoma(PTMC).Methods A total of 104 PTMC patients admitted to our hospital between May 2023 and May 2025 were included in the study.All patients showed no evidence of suspicious lymph node metastasis on preoperative imaging and underwent standard thyroidectomy with central lymph node dissection.Based on postoperative pathological con-firmation of central lymph node metastasis status,patients were classified into an OLNM-positive group(n=53)and an OLNM-negative group(n=51).Baseline characteristics,clinicopathological features,BRAFV600E gene mutation status,and TERT promoter mutation status were compared between the two groups.To identify factors independently associated with OLNM in PTMC patients,multivariate logistic regression analysis was conducted.The area under the receiver operating characteristic curve(AUC)was utilized to assess the predictive performance of a combined model incorporating clinical,pathological,and genetic features for OLNM.Results Compared with the OLNM-negative group,the OLNM-positive group exhibited significantly higher preoperative thyroid-stimulating hormone(TSH)levels(P<0.05).Moreover,the OLNM-positive group demonstrated significantly greater proportions of tumors with diameter>0.5 cm,multifocality,microcalcifications,capsule invasion,extrathyroidal extension,T3 stage,BRAFV600E mutation,and TERT promoter mutation(all P<0.05).Multivariate logistic regression analysis identified preoperative TSH level,tumor diameter>0.5 cm,multifocal lesions,capsule invasion,extrathyroidal extension,T stage,BRAFV600E mutation,and TERT promoter mutation as independent risk factors for OLNM in patients with PTMC(all P<0.05).ROC curve analysis demonstrated that the integrated model combining clinical pathological features-including tumor diameter,number of lesions,microcalcification,capsule invasion,extrathy-roidal extension,and T stage-with genetic markers(BRAFV600E and TERT promoter mutations)exhibited the highest predictive performance,yielding an AUC of 0.940.This was significantly higher than the model based solely on clinical pathological features(AUC=0.736)or those relying exclusively on genetic testing(BRAFV600E:AUC=0.860;TERT:AUC=0.882),with all comparisons reaching statistical significance(P<0.05).Conclusions The integration of clinical pathological features with genetic testing significantly improved the predictive accuracy of OLNM in PTMC patients,surpassing models based solely on individual clinical pathological characteristics or genetic tests alone.This multimodal strategy offers a robust,evidence-based foundation for personalized surgical planning and enhances the precision of clinical decision-making in the management of PTMC.

Objective To conduct an in-depth analysis and feature extraction of the transcriptomics data of 10 types of cancers in order to realize the staging diagnosis of cancer samples.Methods The transcriptomics data of the top 10 cancers having the highest incidence were amassed from the UCSC Xena website,which comprised 4 938 samples and 59 428 genes.With the aid of variational autoencoder,we developed an incremental feature ranking and selection variational autoencoder(IFRSVAE)based on feature importance ranking and incorporating the masking algorithm and the Incremental Feature Selection(IFS).Subsequently,the performance efficiency of our IFRSVAE model was evaluated in conjunction with Random Forest(RF),Support Vector Machine(SVM),and eXtreme Gradient Boosting(XGboost),and it was also compared with other methods.Results Our research extracted 21 features for the ensuing classification.In comparison to the conventional variational autoencoder,recursive feature elimination,and Lasso regression models,the IFRSVAE model attained more favorable performance across all 3 classifiers(highest AUC value,and well performed other indicators).Notably,the IFRSVAE-RF exhibited the most outstanding performance,with an AUC value reaching 85.49%(95%CI:83.24%～87.74%).Moreover,Shapley additive explanations(SHAP)interpretable model illustrated well contributions of the features in our model.Conclusion Our developed IFRSVAE shows certain effectiveness in feature extraction.The constructed IFRSVAE-RF model demonstrates relatively good performance in the task of cancer staging diagnosis,which providing a new and referable idea for research orientation of deep-learning-based diagnostic methods for cancer staging.

Objective As an important monitoring target for gaseous effluents from nuclear facilities, the radioactive inert gas ⁸⁵Kr requires accurate measurement, which is crucial for nuclear safety and radiation protection. This study organized an inter-laboratory comparison activity for ⁸⁵Kr gamma-ray spectrometry measurements, with the aims of identifying potential problems in ⁸⁵Kr measurement and improving detection capabilities and proficiency. Methods Four laboratories participated in this comparison. Each laboratory employed gamma-ray spectrometry to measure the comparison samples, and the results were evaluated using the E_n value. Results The measurement results from all participating institutions were acceptable, with the maximum E_n value being 0.73. The results demonstrate that despite the low gamma-ray branching ratio of ⁸⁵Kr and its susceptibility to interference from the 511 keV positron annihilation peak, gamma-ray spectrometry can achieve reliable activity measurements of ⁸⁵Kr. Conclusion The comparison for ⁸⁵Kr gamma-ray spectrometry tested the capabilities of the participating laboratories in ⁸⁵Kr measurement. It promoted technical exchange and experience sharing among laboratories, laying a foundation for future comparisons involving more complex radioactive inert gases.

Objective:To explore the clinical distribution characteristics and changes of common inhalant allergens and food allergens in all outpatient and inpatient children visiting Shanghai Children′s Hospital from 2020 to 2024, and to provide a basis for the diagnosis, treatment and prevention of allergic diseases in children.Methods:A retrospective cohort study was conducted to retrospectively enroll all outpatient and inpatient children who visited Shanghai Children′s Hospital and underwent serum allergen-specific IgE (sIgE) antibody testing from January 2020 to August 2024, and the characteristics and changing trends of allergens in the past 5 years were analyzed. A total of 127 310 tests were included. There were 76 776 male tests (60.31%) and 50 534 female tests (39.69%). There were 27 392 tests (21.52%) aged 0-3 years (infant group), 51 596 tests (40.53%) aged 4-6 years (preschool group), 44 574 tests (35.01%) aged 7-12 years (school-age group), and 3 748 tests (2.94%) aged 13-18 years (adolescent group). The χ2 test was used for statistical analysis. Results:The difference in total positivity rate between different years was statistically significant ( χ2=2 907.478, P<0.001). The positive rates of inhalant allergens such as house dust, Dermatophagoides pteronyssinus/ Dermatophagoides farinae, cockroach, cat dander, mugwort, Humulus scandens, mold fungi mix, and food allergens such as beef and mutton increased significantly with age (The χ2 values were 649.496, 3 414.686, 303.247, 1 277.408, 40.477, 189.952, 600.737, 203.198, and 15.301, respectively, and the P values were <0.001, <0.001, <0.001, <0.001, <0.001, <0.001,<0.001,<0.001, and 0.002, respectively). The positive rates of inhalant allergen such as Ambrosia elatior (1.59%) and food allergens such as egg white (17.36%), milk (30.48%), shrimp (8.27%), crab (8.13%), codfish (2.61%), salmon (0.66%), mytilus edulis (2.89%), lobster/scallop (5.27%), cashew nuts (5.09%), peanuts (3.54%), and soybean (1.73%) were highest at the age of 0-3 years and decreased significantly with age (The χ2 values were 10.365, 2 407.443, 139.085, 872.548, 870.245, 106.823, 47.674, 47.244, 559.422, 369.800, 384.788, 153.660, respectively, and the P values were 0.016, 0.000,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001,<0.001, respectively). Inhaled allergens mainly have a greater impact on children with respiratory-related diseases such as allergic rhinitis and asthma, while food allergens mainly have a greater impact on children with atopic dermatitis/eczema. The positive rate of sIgE of various allergens in the allergic rhinitis combined asthma group were higher than that of allergic rhinitis alone, and the sIgE positive rate of total allergens and inhaled allergens was significantly higher than that of allergic rhinitis alone ( χ2=20.851, 39.155, the P values were both<0.001). Among them, the sIgE positive rate of Ambrosia elatior and cashew nuts showed significant difference ( χ2=5.044, 8.420, P=0.025, 0.004); and the sIgE positive rate of Dermatophagoides pteronyssinus/ Dermatophagoides farinae, cat dander, grass pollens mix and mold fungi mix had extremely significant difference ( χ2=26.409, 25.990, 21.283, 16.411, the P values were all <0.001). The inhaled allergens and food allergens with the highest positive rates in the 5 years were Dermatophagoides pteronyssinus/ Dermatophagoides farinae (56.21%) and milk (47.47%), and as time went by, the positive rates gradually decreased. There is a moderate correlation between the three allergens of Ambrosia elatior, Amaranthus retroflexus, and tree pollens mix (0.55, 0.70, 0.63), and there is a moderate correlation between mango and tree pollens mix (0.50). Conclusion:Dermatophagoides pteronyssinus/ Dermatophagoides farinae, cat dander, dog dander, egg white, and milk may be important allergens for children in Shanghai City from 2020 to 2024. The positive rates vary among different genders, age groups, and disease groups, but the positive rates of Dermatophagoides pteronyssinus/ Dermatophagoides farinae, milk and cat dander allergens remain in the top three.

Gongfa is an essential approach to prevent and treat diseases in traditional Chinese medicine(TCM),often used to prevent a disease before it arises.Guided by TCM and modern scientific theories,the Tuina(Chinese therapeutic massage)Gongfa prescription theory implements the principle,method,prescription,and form in clinical pattern-identified treatment to prescribe the corresponding Tuina Gongfa prescription,i.e.,to prescribe a basic Tuina Gongfa prescription,specifically for a systemic disease,and modify Gongfa forms based on the basic prescription according to different patterns.The Gongfa prescription for pulmonary diseases designs corresponding Gongfa forms from six perspectives:lifting Yang,securing the exterior,opening the orifices,soothing the chest,harmonizing the stomach,and regulating Qi to prevent and treat diseases.The application of the pulmonary Gongfa prescription indicates the potential to apply the Tuina Gongfa prescription theory for the clinical prevention,treatment,and rehabilitation of disorders of other systems,thereby fully realizing the unique role of TCM Gongfa.