1.Explainable Machine Learning Model for Predicting Prognosis in Patients with Malignant Tumors Complicated by Acute Respiratory Failure: Based on the eICU Collaborative Research Database in the United States
Zihan NAN ; Linan HAN ; Suwei LI ; Ziyi ZHU ; Qinqin ZHU ; Yan DUAN ; Xiaoting WANG ; Lixia LIU
Medical Journal of Peking Union Medical College Hospital 2026;17(1):98-108
To develop and validate a model for predicting intensive care unit (ICU) mortality risk in patients with malignant tumors complicated by acute respiratory failure (ARF) based on an explainable machine learning framework. Clinical data of patients with malignant tumors and ARF were extracted from the eICU Collaborative Research Database in the United States, including demographic characteristics, comorbidities, vital signs, laboratory test indicators, and major interventions within the first 24 hours after ICU admission.The study outcome was ICU death.Enrolled patients were randomly divided into a training set and a validation set at a ratio of 7:3.Predictor variables were selected using least absolute shrinkage and selection operator (LASSO) regression.Five machine learning algorithms-extreme gradient boosting (XGBoost), support vector machine (SVM), Logistic regression, multilayer perceptron (MLP), and C5.0 Decision Tree-were employed to construct predictive models.Model performance was evaluated based on the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and other metrics.The optimal model was further interpreted using the Shapley additive explanations (SHAP) algorithm. A total of 3196 patients with malignant tumors complicated by ARF were included.The training set comprised 2, 261 patients and the validation set 935 patients; 683 patients died during ICU stay, while 2513 survived.LASSO regression ultimately selected 12 variables closely associated with patient ICU outcomes, including sepsis comorbidity, use of vasoactive drugs, and within the first 24 hours after ICU admission: minimum mean arterial pressure, maximum heart rate, maximum respiratory rate, minimum oxygen saturation, minimum serum bicarbonate, minimum blood urea nitrogen, maximum white blood cell count, maximum mean corpuscular volume, maximum serum potassium, and maximum blood glucose.After model evaluation, the XGBoost model demonstrated the best performance.The AUCs for predicting ICU mortality risk in the training and validation sets were 0.940 and 0.763, respectively; accuracy was 88.3% and 81.2%;sensitivity was 98.5% and 95.9%.Its predictive performance also remained optimal in sensitivity analyses.SHAP analysis indicated that the top five variables contributing to the model's predictions were minimum oxygen saturation, minimum serum bicarbonate, minimum mean arterial pressure, use of vasoactive drugs, and maximum white blood cell count. This study successfully developed a mortality risk prediction model for ICU patients with malignant tumors complicated by ARF based on a large-scale dataset and performed explainability analysis.The model aids clinicians in early identification of high-risk patients and implementing individualized interventions.
2.Intervention Strategies for Heart Failure with Preserved Ejection Fraction Using Combined Classical Formulas Based on the Theory of "Disease of Both Blood and Water"
Yuzhi JIA ; Qingyong HE ; Jie WANG ; Xin ZHAO ; Ziyi WANG ; Dongmei LI ; Junqiao AN
Journal of Traditional Chinese Medicine 2026;67(4):370-374
Based on the theory of "disease of both blood and water" in Essentials from the Golden Cabinet (《金匮要略》), and in combination with the dynamic syndrome evolution of heart failure with preserved ejection fraction (HFpEF), this paper systematically clarifies the pathomechanism of HFpEF, characterized by yang deficiency as the root, blood stasis as the pivotal factor and water retention as the manifestation. Accordingly, the therapeutic principles have been proposed, which are warming yang and banking up original qi to consolidate the root, activating blood and unblocking collaterals to smooth the mechanism, and promoting urination and regulating pivot to remove the branch. On this basis, a compound formula structure of "one monarch, one minister and one assistant" is established, forming an integrated intervention strategy that synergistically combines the three methods of warming yang, activating blood, and promoting urination through combined classical formulas. Zhenwu Decoction (真武汤), which warms yang and dissolves rheum, is used to consolidate the root and directly target the source of yang deficiency, serving as the monarch; Guizhi Fuling Pills (桂枝茯苓丸), which activates blood, promotes urination and unblocks the pivot, assists in interrupting the binding of blood stasis and water retention, serving as the minister; Tingli Dazao Xiefei Decoction (葶苈大枣泻肺汤), which regulates qi, disperses retained fluids, and eliminates the manifestation, alleviates acute water-retention symptoms, serving as the assistant. This compound formula is warming without being drying, diuretic without being drastic, and dispels stasis without consuming blood, thereby achieving the therapeutic effects of warming yang, activating blood, and promoting urination.
3.Association of outdoor activity level and myopia among children and adolescents in Shanghai
Chinese Journal of School Health 2025;46(1):18-23
Objective:
To analyze the status of outdoor activities on weekends among children and adolescents of different educational stages in Shanghai and their impact on myopia, so as to provide a basis for formulating more specific prevention and control protocol of myopia.
Methods:
From September to October 2022, a stratified cluster random sampling method was employed to select 84 schools (27 kindergartens, 21 primary schools, 15 junior high schools and 21 high schools) across Shanghai, enrolling a total of 28 654 children and adolescents aged 4 to 18 for the study. Ophthalmic examinations were conducted to ascertain the prevalence of myopia among children and adolescents. Additionally, a questionnaire survey was administered to collect data on outdoor activity duration and associated factors. Multivariate Logistic regression analysis was utilized to investigate the associated factors of outdoor activity levels on weekends.
Results:
The overall myopia detection rate among children and adolescents was 58.4%, with a higher rate observed in girls (59.2%) compared to boys (57.6%). The myopia detection rates for children and adolescents with an average daily outdoor activity duration of ≥2 h and <2 h on weekends were 54.6% and 68.8%, and the differences were statistically significant ( χ 2=8.12,460.89, P <0.01). Multivariable Logistic regression analysis revealed that girls ( OR =0.80), those with a myopic parent ( OR =0.68), schools from urban districts ( OR =0.72), higher education stages (primary school: OR =0.65, junior high school: OR =0.24, high school: OR =0.14) and spending≥2 h/d on homework during weekends ( OR =0.57) among children and adolescents were less likely to engage in outdoor activities for ≥2 h on weekends ( P <0.01). After incorporating gender, parental myopia status, educational stage, school location, average daily duration on weekends for spending on homework, electronic product usage and outdoor activities as dependent variables in a multivariate Logistic regression analysis, the results showed that children and adolescents with an average outdoor activity duration for ≥2 h on weekends had a lower risk of myopia ( OR =0.86, P < 0.01).
Conclusions
The level of outdoor activity among children and adolescents on weekends needs to be improved. Outdoor activities on weekends is an associated factor for myopia among children and adolescents. Particularly, girls, those with myopic parents, schools from urban districts, and spending long hours on homework during weekends among children and adolescents require increased attention.
4.Visual acuity and corrected visual acuity of children and adolescents in Shanghai City
Chinese Journal of School Health 2025;46(1):24-28
Objective:
To investigate the visual acuity and correction conditions of children and adolescents in Shanghai, so as to provide a scientific basis for developing intervention measures to prevent myopia and protect vision among children and adolescents.
Methods:
From October to December 2022, a stratified cluster random sampling survey was conducted, involving 47 034 students from 16 municipal districts in Shanghai, covering kindergartens (≥5 years), primary schools, middle schools, general high schools and vocational high schools. According to the Guidelines for Screening Refractive Errors in Primary and Secondary School Students, the Standard Logarithmic Visual acuity Chart was used to examine naked vision and corrected vision of students, and general information was collected. The distribution and severity of visual impairment in different age groups were analyzed, and χ 2 tests and multivariate Logistic regression were used to explore factors associated with visual impairment.
Results:
The detection rate of visual impairment among children and adolescents was 76.2%, with a higher rate among females (78.8%) than males ( 73.8 %), higher among Han ethic students ( 76.2 %) than minority students (71.2%), and higher among urban students (76.7%) than suburban students (75.8%), all with statistically significant differences ( χ 2=162.6, 10.4, 5.5, P <0.05). The rate of visual impairment initially decreased and then increased with age, reaching its lowest at age 7 (53.8%) and peaking at age 17 (89.6%) ( χ 2 trend = 3 467.0 , P <0.05). Severe visual impairment accounted for the majority, at 56.6%, and there was a positive correlation between the severity of visual impairment and age among children and adolescents ( r =0.45, P <0.05). Multivariate Logistic regression showed that age, BMI, gender, ethnicity and urban suburban status were associated with visual impairment ( OR =1.18, 1.01, 1.38 , 0.79, 0.88, P <0.05). Among those with moderate to severe visual impairment, the rate of spectacle lens usage was 62.8%, yet only 44.8 % of those who used spectacle lens had fully corrected visual acuity. Females (64.9%) had higher spectacle lens usage rates than males (60.6%), and general high school students had the highest spectacle lens usage (83.9%), and there were statistically significant differences in gender and academic stages ( χ 2=57.7, 4 592.8, P <0.05).
Conclusions
The rate of spectacle lens usage among students with moderate to severe visual impairment is relatively low, and even after using spectacle lens, some students still do not achieve adequate corrected visual acuity. Efforts should focus on enhancing public awareness of eye health and refractive correction and improving the accessibility of related health services.
5.Metabolism and metabolomics in senescence, aging, and age-related diseases: a multiscale perspective.
Ziyi WANG ; Hongying ZHU ; Wei XIONG
Frontiers of Medicine 2025;19(2):200-225
The pursuit of healthy aging has long rendered aging and senescence captivating. Age-related ailments, such as cardiovascular diseases, diabetes, and neurodegenerative disorders, pose significant threats to individuals. Recent studies have shed light on the intricate mechanisms encompassing genetics, epigenetics, transcriptomics, and metabolomics in the processes of senescence and aging, as well as the establishment of age-related pathologies. Amidst these underlying mechanisms governing aging and related pathology metabolism assumes a pivotal role that holds promise for intervention and therapeutics. The advancements in metabolomics techniques and analysis methods have significantly propelled the study of senescence and aging, particularly with the aid of multiscale metabolomics which has facilitated the discovery of metabolic markers and therapeutic potentials. This review provides an overview of senescence and aging, emphasizing the crucial role metabolism plays in the aging process as well as age-related diseases.
Humans
;
Aging/metabolism*
;
Metabolomics/methods*
;
Neurodegenerative Diseases/metabolism*
;
Cardiovascular Diseases/metabolism*
6.Elucidating the role of artificial intelligence in drug development from the perspective of drug-target interactions.
Boyang WANG ; Tingyu ZHANG ; Qingyuan LIU ; Chayanis SUTCHARITCHAN ; Ziyi ZHOU ; Dingfan ZHANG ; Shao LI
Journal of Pharmaceutical Analysis 2025;15(3):101144-101144
Drug development remains a critical issue in the field of biomedicine. With the rapid advancement of information technologies such as artificial intelligence (AI) and the advent of the big data era, AI-assisted drug development has become a new trend, particularly in predicting drug-target associations. To address the challenge of drug-target prediction, AI-driven models have emerged as powerful tools, offering innovative solutions by effectively extracting features from complex biological data, accurately modeling molecular interactions, and precisely predicting potential drug-target outcomes. Traditional machine learning (ML), network-based, and advanced deep learning architectures such as convolutional neural networks (CNNs), graph convolutional networks (GCNs), and transformers play a pivotal role. This review systematically compiles and evaluates AI algorithms for drug- and drug combination-target predictions, highlighting their theoretical frameworks, strengths, and limitations. CNNs effectively identify spatial patterns and molecular features critical for drug-target interactions. GCNs provide deep insights into molecular interactions via relational data, whereas transformers increase prediction accuracy by capturing complex dependencies within biological sequences. Network-based models offer a systematic perspective by integrating diverse data sources, and traditional ML efficiently handles large datasets to improve overall predictive accuracy. Collectively, these AI-driven methods are transforming drug-target predictions and advancing the development of personalized therapy. This review summarizes the application of AI in drug development, particularly in drug-target prediction, and offers recommendations on models and algorithms for researchers engaged in biomedical research. It also provides typical cases to better illustrate how AI can further accelerate development in the fields of biomedicine and drug discovery.
7.Spatial metabolomics combined with machine learning in colon cancer diagnosis research.
Ling WENG ; Huanhuan WANG ; Chunxiang ZHAI ; Qi WANG ; Yanyan GUO ; Ziyi ZHONG ; Chenying MA ; Jing WANG
Journal of Pharmaceutical Analysis 2025;15(8):101367-101367
Image 1.
8.Textual Research on Classical Formula Mulisan
Dongsen HU ; Xiangyang ZHANG ; Canran XIE ; Jiawei SHI ; Ziyi WANG ; Zhuoyan ZHOU ; Lin ZHANG ; Yexin CHEN
Chinese Journal of Experimental Traditional Medical Formulae 2025;31(7):191-200
The classic formula Mulisan is the 45th of the 93 formulas in the Catalogue of Ancient Classic Formulas (second batch) of Han medicine published by the National Administration of Traditional Chinese Medicine. It consists of Ostreae Concha, Astragali Radix, Ephedrae Radix et Rhizoma, and wheat, with the effect of replenishing qi and stopping sweating. It is a common formula in the clinical treatment with traditional Chinese medicine. This study analyzes the historical evolution, composition, dosage, original plants and their processing methods, decocting method, efficacy, indications, and modern clinical application of Mulisan by tracing, comparative analysis, and bibliometric methods. The results showed that Mulisan firstly appeared in the Pulse Classic written by WANG Shuhe in the Western Jin Dynasty. The formulation idea can be traced back to the Important Prescriptions Worth a Thousand Gold for Emergency in the Tang Dynasty. The herb composition, dosage, efficacy, and indications of Mulisan were first recorded in the Treatise on Diseases, Patterns, and formulas Related to Unification of the Three Etiologies in the Southern Song dynasty. In terms of original plants and their processing methods, Ostreae Concha is the shell of Ostrea rivularis, which should be calcined before use. Astragali Radix and Ephedrae Radix et Rhizoma are the dried roots of Astragalus membranaceus var. mongholicus and Ephedra sinica, respectively, the raw material of which should be used. Wheat is the dried mature fruit of T. aestivum, which can be used without processing, while the stir-fried fruit, being thin and deflated, demonstrates better effect. The composition of Mulisan is Ostreae Concha 8.26 g, Astragali Radix 8.26 g, Ephedrae Radix et Rhizoma 8.26 g, and wheat 7.92 g. The medicinal materials should be ground into coarse powder and decocted with 450 mL water to reach a volume of 240 mL, and the decoction should be taken warm. In modern clinical practice, Mulisan has a wide range of indications, including spontaneous sweating and night sweating caused by Yang deficiency or Qi deficiency. The clinical disease spectrum treated by Mulisan involves endocrine system diseases, neurological diseases, respiratory system diseases, and cancer. This formula plays a significant role in the treatment of internal medicine diseases in traditional Chinese medicine. This study aims to provide a scientific basis for the subsequent research, development, and clinical application of Mulisan.
9.Characterization of Animal Models of Type 2 Diabetes Mellitus with Hypertension Based on Clinical Symptoms in Traditional Chinese and Western Medicine
Ziyi WANG ; Xingyu FAN ; Yuzhi JIA ; Qingyong HE
Chinese Journal of Experimental Traditional Medical Formulae 2025;31(7):230-236
Type 2 diabetes mellitus (T2DM) and hypertension are common and frequent chronic non-communicable diseases, which often coexist in clinical practice, resulting in a large number of cardiovascular events and deaths, and their case fatality rate far exceeds that of other factors such as dyslipidemia and obesity. Based on the diagnostic standards, guidelines, and animal model evaluation methods of T2DM with hypertension at home and abroad, this study summarized, evaluated, and analyzed the characteristics of the existing animal models of T2DM with hypertension based on the clinical symptoms in traditional Chinese and Western medicine. The animal models showing high fitting degrees with the clinical symptoms in Western medicine are mainly established by injection of streptozocin (STZ) in SHR rats in the surgical induction and chemical induction methods and feeding a high-fat and high-salt diet combined with STZ injection in SD rats in the dietary induction methods. The models showcasing high fitting degrees with the clinical symptoms in traditional Chinese medicine (TCM) are mainly established by the surgical induction method. Considering the fitting degrees and the advantages and disadvantages, the ideal modeling method for T2DM with hypertension is the two-kidney, one clip (2K1C) method (a surgical induction method) combined with feeding a high-fat and high-sugar diet and STZ injection. However, the available models lack the characteristics of TCM and the evaluation indicators have poor specificity. This study found that there are few animal models of T2DM with hypertension considering the characteristics of both disease and syndrome, which may be related to the identification and attribution of TCM syndromes in animal macroscopic information. In view of this problem, it is suggested that the evaluation criteria should be established and improved for the animal models combining disease and syndrome, which can help to evaluate the fitting degree of the pathological characteristics of different syndromes in the animal models of T2DM with hypertension. In this way, ideal animal models of T2DM with hypertension can be established to simulate the disease occurrence and development in the human body. The animal models are expected to provide an ideal approach for the further research on the pathogenesis of T2DM and its prevention and treatment with TCM, which is of great significance for the treatment and prevention of T2DM with hypertension and the prognosis of its complications. At the same time, breakthroughs in the basic syndrome models of comorbidities are expected to lay a foundation for the leapfrog development of TCM research.
10.Textual Research on Classical Formula Mulisan
Dongsen HU ; Xiangyang ZHANG ; Canran XIE ; Jiawei SHI ; Ziyi WANG ; Zhuoyan ZHOU ; Lin ZHANG ; Yexin CHEN
Chinese Journal of Experimental Traditional Medical Formulae 2025;31(7):191-200
The classic formula Mulisan is the 45th of the 93 formulas in the Catalogue of Ancient Classic Formulas (second batch) of Han medicine published by the National Administration of Traditional Chinese Medicine. It consists of Ostreae Concha, Astragali Radix, Ephedrae Radix et Rhizoma, and wheat, with the effect of replenishing qi and stopping sweating. It is a common formula in the clinical treatment with traditional Chinese medicine. This study analyzes the historical evolution, composition, dosage, original plants and their processing methods, decocting method, efficacy, indications, and modern clinical application of Mulisan by tracing, comparative analysis, and bibliometric methods. The results showed that Mulisan firstly appeared in the Pulse Classic written by WANG Shuhe in the Western Jin Dynasty. The formulation idea can be traced back to the Important Prescriptions Worth a Thousand Gold for Emergency in the Tang Dynasty. The herb composition, dosage, efficacy, and indications of Mulisan were first recorded in the Treatise on Diseases, Patterns, and formulas Related to Unification of the Three Etiologies in the Southern Song dynasty. In terms of original plants and their processing methods, Ostreae Concha is the shell of Ostrea rivularis, which should be calcined before use. Astragali Radix and Ephedrae Radix et Rhizoma are the dried roots of Astragalus membranaceus var. mongholicus and Ephedra sinica, respectively, the raw material of which should be used. Wheat is the dried mature fruit of T. aestivum, which can be used without processing, while the stir-fried fruit, being thin and deflated, demonstrates better effect. The composition of Mulisan is Ostreae Concha 8.26 g, Astragali Radix 8.26 g, Ephedrae Radix et Rhizoma 8.26 g, and wheat 7.92 g. The medicinal materials should be ground into coarse powder and decocted with 450 mL water to reach a volume of 240 mL, and the decoction should be taken warm. In modern clinical practice, Mulisan has a wide range of indications, including spontaneous sweating and night sweating caused by Yang deficiency or Qi deficiency. The clinical disease spectrum treated by Mulisan involves endocrine system diseases, neurological diseases, respiratory system diseases, and cancer. This formula plays a significant role in the treatment of internal medicine diseases in traditional Chinese medicine. This study aims to provide a scientific basis for the subsequent research, development, and clinical application of Mulisan.


Result Analysis
Print
Save
E-mail