BACKGROUND: Postoperative pulmonary complications (PPCs) are major adverse events in neurosurgical patients. This study aimed to develop and validate machine learning models predicting PPCs after neurosurgery. METHODS: PPCs were defined according to the European Perioperative Clinical Outcome standards as occurring within 7 postoperative days. Data of cases meeting inclusion/exclusion criteria were extracted from the anesthesia information management system to create three datasets: The development (data of Huashan Hospital, Fudan University from 2018 to 2020), temporal validation (data of Huashan Hospital, Fudan University in 2021) and external validation (data of other three hospitals in 2023) datasets. Machine learning models of six algorithms were trained using either 35 retrievable and plausible features or the 11 features selected by Lasso regression. Temporal validation was conducted for all models and the 11-feature models were also externally validated. Independent risk factors were identified and feature importance in top models was analyzed. RESULTS: PPCs occurred in 712 of 7533 (9.5%), 258 of 2824 (9.1%), and 207 of 2300 (9.0%) patients in the development, temporal validation and external validation datasets, respectively. During cross-validation training, all models except Bayes demonstrated good discrimination with an area under the receiver operating characteristic curve (AUC) of 0.840. In temporal validation of full-feature models, deep neural network (DNN) performed the best with an AUC of 0.835 (95% confidence interval [CI]: 0.805-0.858) and a Brier score of 0.069, followed by Logistic regression (LR), random forest and XGBoost. The 11-feature models performed comparable to full-feature models with very close but statistically significantly lower AUCs, with the top models of DNN and LR in temporal and external validations. An 11-feature nomogram was drawn based on the LR algorithm and it outperformed the minimally modified Assess respiratory RIsk in Surgical patients in CATalonia (ARISCAT) and Laparoscopic Surgery Video Educational Guidelines (LAS VEGAS) scores with a higher AUC (LR: 0.824, ARISCAT: 0.672, LAS: 0.663). Independent risk factors based on multivariate LR mostly overlapped with Lasso-selected features, but lacked consistency with the important features using the Shapley additive explanation (SHAP) method of the LR model. CONCLUSIONS: The developed models, especially the DNN model and the nomogram, had good discrimination and calibration, and could be used for predicting PPCs in neurosurgical patients. The establishment of machine learning models and the ascertainment of risk factors might assist clinical decision support for improving surgical outcomes. TRIAL REGISTRATION ChiCTR 2100047474; https://www.chictr.org.cn/showproj.html?proj=128279 .
Adult ; Aged ; Female ; Humans ; Male ; Middle Aged ; Algorithms ; Lung Diseases/etiology* ; Machine Learning ; Neurosurgical Procedures/adverse effects* ; Postoperative Complications/diagnosis* ; Risk Factors ; ROC Curve

BACKGROUND: Non-invasive computed tomography angiography (CTA)-based fractional flow reserve (CT-FFR) could become a gatekeeper to invasive coronary angiography. Deep learning (DL)-based CT-FFR has shown promise when compared to invasive FFR. To evaluate the performance of a DL-based CT-FFR technique, DeepVessel FFR (DVFFR). METHODS: This retrospective study was designed for iScheMia Assessment based on a Retrospective, single-center Trial of CT-FFR (SMART). Patients suspected of stable coronary artery disease (CAD) and undergoing both CTA and invasive FFR examinations were consecutively selected from the Beijing Anzhen Hospital between January 1, 2016 to December 30, 2018. FFR obtained during invasive coronary angiography was used as the reference standard. DVFFR was calculated blindly using a DL-based CT-FFR approach that utilized the complete tree structure of the coronary arteries. RESULTS: Three hundred and thirty nine patients (60.5 ±10.0 years and 209 men) and 414 vessels with direct invasive FFR were included in the analysis. At per-vessel level, sensitivity, specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) of DVFFR were 94.7%, 88.6%, 90.8%, 82.7%, and 96.7%, respectively. The area under the receiver operating characteristics curve (AUC) was 0.95 for DVFFR and 0.56 for CTA-based assessment with a significant difference (P < 0.0001). At patient level, sensitivity, specificity, accuracy, PPV and NPV of DVFFR were 93.8%, 88.0%, 90.3%, 83.0%, and 95.8%, respectively. The computation for DVFFR was fast with the average time of 22.5 ± 1.9 s. CONCLUSIONS The results demonstrate that DVFFR was able to evaluate lesion hemodynamic significance accurately and effectively with improved diagnostic performance over CTA alone. Coronary artery disease (CAD) is a critical disease in which coronary artery luminal narrowing may result in myocardial ischemia. Early and effective assessment of myocardial ischemia is essential for optimal treatment planning so as to improve the quality of life and reduce medical costs.

Patient 1,a 12-day-old female infant,presented with fever,cough,dyspnea,and elevated infection markers,requiring respiratory support.Metagenomic next-generation sequencing(mNGS)of blood and bronchoalveolar lavage fluid revealed Legionella pneumophila(LP),leading to diagnoses of LP pneumonia and LP sepsis.The patient was treated with erythromycin for 15 days and azithromycin for 5 days,resulting in recovery and discharge.Patient 2,an 11-day-old female infant,presented with dyspnea,fever,elevated infection markers,and multiple organ dysfunction,requiring mechanical ventilation.mNGS of blood and cerebrospinal fluid indicated LP,leading to diagnoses of LP pneumonia,LP sepsis,and LP intracranial infection.The patient was treated with erythromycin for 19 days and was discharged after recovery.Neonatal LP pneumonia lacks specific clinical symptoms,and azithromycin is the preferred antimicrobial agent.The use of mNGS can provide early and definitive diagnosis for severe neonatal pneumonia of unknown origin.

Objective To explore early postoperative gait characteristics and clinical outcomes after total knee arthroplasty(TKA).Methods From February 2023 to July 2023,26 patients with unilateral knee osteoarthritis(KOA)were treated with TKA,including 4 males and 22 females,aged from 57 to 85 years old with an average of(67.58±6.49)years old;body mass in-dex(BMI)ranged from 18.83 to 38.28 kg·m-2 with an average of(26.43±4.15)kg·m-2;14 patients on the left side,12 pa-tients on the right side;according to Kellgren-Lawrence(K-L)classification,6 patients with grade Ⅲ and 20 patients with grade Ⅳ;the courses of disease ranged from 1 to 14 years with an average of(5.54±3.29)years.Images and videos of standing up and walking,walking side shot,squatting and supine kneeling were taken with smart phones before operation and 6 weeks after operation.The human posture estimation framework OpenPose were used to analyze stride frequency,step length,step length,step speed,active knee knee bending angle,stride length,double support phase time,as well as maximum hip flexion angle and maximum knee bending angle on squatting position.Western Ontario and McMaster Universities(WOMAC)arthritis index and Knee Society Score(KSS)were used to evaluate clinical efficacy of knee joint.Results All patients were followed up for 5 to 7 weeks with an average of(6.00±0.57)weeks.The total score of WOMAC decreased from(64.85±11.54)before op-eration to(45.81±7.91)at 6 weeks after operation(P＜0.001).The total KSS was increased from(101.19±9.58)before opera-tion to(125.50±10.32)at 6 weeks after operation(P＜0.001).The gait speed,stride frequency and stride length of the affected side before operation were(0.32±0.10)m·s-1,(96.35±24.18)steps·min-1,(0.72±0.14)m,respectively;and increased to(0.48±0.11)m·s 1,(104.20±22.53)steps·min-1,(0.79±0.10)m at 6 weeks after operation(P＜0.05).The lower limb support time and active knee bending angle decreased from(0.31±0.38)sand(125.21±11.64)° before operation to(0.11±0.04)s and(120.01±13.35)° at 6 weeks after operation(P＜0.05).Eleven patients could able to complete squat before operation,13 patients could able to complete at 6 weeks after operation,and 9 patients could able to complete both before operation and 6 weeks after operation.In 9 patients,the maximum bending angle of crouching position was increased from 76.29° to 124.11° before operation to 91.35° to 134.12° at 6 weeks after operation,and the maximum bending angle of hip was increased from 103.70° to 147.25° before operation to 118.61° to 149.48° at 6 weeks after operation.Conclusion Gait analysis technology based on artificial intelligence image recognition is a safe and effective method to quantitatively identify the changes of pa-tients'gait.Knee pain of KOA was relieved and the function was improved,the supporting ability of the affected limb was im-proved after TKA,and the patient's stride frequency,stride length and stride speed were improved,and the overall movement rhythm of both lower limbs are more coordinated.

As artificial intelligence technology rapidly advances, its deployment within the medical sector presents substantial ethical challenges. Consequently, it becomes crucial to create a standardized, transparent, and secure framework for processing medical data. This includes setting the ethical boundaries for medical artificial intelligence and safeguarding both patient rights and data integrity. This consensus governs every facet of medical data handling through artificial intelligence, encompassing data gathering, processing, storage, transmission, utilization, and sharing. Its purpose is to ensure the management of medical data adheres to ethical standards and legal requirements, while safeguarding patient privacy and data security. Concurrently, the principles of compliance with the law, patient privacy respect, patient interest protection, and safety and reliability are underscored. Key issues such as informed consent, data usage, intellectual property protection, conflict of interest, and benefit sharing are examined in depth. The enactment of this expert consensus is intended to foster the profound integration and sustainable advancement of artificial intelligence within the medical domain, while simultaneously ensuring that artificial intelligence adheres strictly to the relevant ethical norms and legal frameworks during the processing of medical data.
Artificial Intelligence/legislation & jurisprudence* ; Humans ; Consensus ; Computer Security/standards* ; Confidentiality/ethics* ; Informed Consent/ethics*

Ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry(UPLC-Q-TOF-MS/MS) was used to rapidly identify the chemical components in Dracocephalum moldavica, and UPLC was employed to determine the content of its main components. MS analysis was performed using an electrospray ionization(ESI) source and data were collected in the negative ion mode. By comparing the retention time and mass spectra of reference compounds, and using a self-built compound database and the PubChem database, 68 compounds were identified from D. moldavica, including 36 flavonoids, 22 phenylpropanoids, 4 phenols, and 6 other compounds. On this basis, a UPLC quantitative method was established to simultaneously determine 8 main components, i.e., luteolin-7-O-glucuronide, apigenin-7-O-glucuronide, rosmarinic acid, diosmetin-7-O-glucuronide, tilianin, acacetin-7-O-glucuronide, acacetin-7-O-(6″-O-malonyl)-glucoside, and acacetin. A Waters ACQUITY BEH C_(18) column(2.1 mm × 100 mm, 1.7 μm) was used, with acetonitrile and a water solution containing 0.1% formic acid and 0.1% phosphoric acid as the mobile phase for gradient elution. The detection wavelength was set at 330 nm, with a flow rate of 0.4 mL·min~(-1), and the column temperature was maintained at 35 ℃. The 8 components demonstrated good linearity(r≥0.999 9) over a wide mass concentration range(50 or 100 times). The average recovery rate ranged from 97.5% to 105.1%, and the relative standard deviations(RSDs) were 0.90% to 3.4%(n= 6), indicating that the method was simple, accurate, and reliable. In 17 batches of D. moldavica samples, the content of these 8 components ranged from 0.405 to 2.10, 0.063 to 0.342, 0.446 to 2.43, 0.415 to 1.47, 1.57 to 4.34, 0.173 to 0.386, 1.00 to 5.40, and 0.069 to 0.207 mg·g~(-1), respectively. These results indicate significant differences in the internal quality of the samples, highlighting the need for strict quality control to ensure their pharmacodynamic efficacy. This study provides a scientific basis for the rapid discovery of pharmacodynamic substances, comprehensive quality control, and the formulation or revision of quality standards for D. moldavica.
Tandem Mass Spectrometry/methods* ; Chromatography, High Pressure Liquid/methods* ; Drugs, Chinese Herbal/chemistry* ; Lamiaceae/chemistry* ; Flavonoids/chemistry*

Objective: To analyze the status quo of the knowledge and related factors of cancer prevention and treatment among residents in Liaoning Province in 2021. Methods: From August to November 2021, through network sampling method, 17 474 permanent residents aged 15-69 years in Liaoning Province were surveyed. The WeChat public account was used to collect information such as demographic characteristics and core knowledge of cancer prevention and treatment. The Chi-square test was used to compare the difference of the level of the cancer prevention and treatment knowledge among different groups. The multivariate logistic regression model was used to analyze the related factors. Results: Among the 17 474 subjects, 43.1% (7 528) were male and 58.7% (10 262) were urban residents. The overall awareness rate was 72.3%, and the awareness rate of cancer cognition, prevention, early diagnosis and treatment, cancer management and rehabilitation were 71.4%, 67.6%, 72.7%, 83.4% and 63.5%, respectively. The multivariate logistic regression model showed that the residents who were man (OR: 0.850, 95%CI: 0.781-0.925), in rural areas (OR: 0.753, 95%CI: 0.694-0.817), 55-59 years old (OR: 0.851, 95%CI: 0.751-0.963), quitters (OR: 0.721, 95%CI: 0.640-0.813) and smoker (OR: 0.724, 95%CI: 0.654-0.801) had lower awareness rates, while the residents who were 35-54 years old (OR: 1.312, 95%CI: 1.202-1.432), with an educational level of junior high school/senior high school/college degree or above (OR: 1.834-5.130, 95%CI: 1.575-6.047), technical personnel (OR: 1.592, 95%CI: 1.367-1.854), civil servant/institution staff (OR: 1.282, 95%CI: 1.094-1.503), enterprise/business/service staff (OR: 1.218, 95%CI: 1.071-1.385), retired (OR: 1.324, 95%CI: 1.114-1.573) and with family history of cancer (OR: 1.369, 95%CI: 1.266-1.481) had higher awareness rates. Conclusion: The level of the awareness of core knowledge of cancer prevention and treatment among residents in Liaoning Province has met the requirements of the Healthy China Action. Region, gender, education level, age, family history of cancer and smoking are relevant factors.
Adult ; Female ; Humans ; Male ; Middle Aged ; China ; Health Knowledge, Attitudes, Practice ; Neoplasms/prevention & control* ; Surveys and Questionnaires ; Adolescent ; Young Adult ; Aged

Chronic kidney disease (CKD) is an important global public health problem that greatly threatens population health. Application of risk prediction model is a crucial way for the primary prevention of CKD, which can stratify the risk for developing CKD and identify high-risk individuals for more intensive interventions. By now, more than twenty risk prediction models for CKD have been developed worldwide. There are also four domestic risk prediction models developed for Chinese population. However, none of these models have been recommended in clinical guidelines yet. The existing risk prediction models have some limitations in terms of outcome definition, predictors, strategies for handling missing data, and model derivation. In the future, the applications of emerging biomarkers and polygenic risk scores as well as advances in machine learning methods will provide more possibilities for the further improvement of the model.
Humans ; Renal Insufficiency, Chronic ; Risk Factors ; Biomarkers

Purpose: This study aimed to evaluate the diagnostic value of combined fine-needle aspiration (FNA) with core needle biopsy (CNB) in thyroid nodules. Methods: FNA and CNB were performed simultaneously on 703 nodules. We compared the proportions of inconclusive results and the diagnostic performance for malignancy among FNA, CNB, and combined FNA/CNB for different nodule sizes. Results: Combined FNA/CNB showed lower proportions of inconclusive results than CNB for all nodules (2.8% vs. 5.7%, P<0.001), nodules ≤1.0 cm (4.9% vs. 7.3%, P=0.063), nodules >1.0 cm (2.0% vs. 5.0 %, P<0.001), nodules ≤1.5 cm (3.8% vs. 7.9 %, P<0.001), and nodules >1.5 cm (2.1% vs. 3.9 %, P=0.016). The sensitivity of combined FNA/CNB in predicting malignancy was significantly higher than that of CNB (89.0% vs. 80.0%, P<0.001) and FNA (89.0% vs. 58.1%, P<0.001) for all nodules. Within American College of Radiology Thyroid and Imaging Reporting and Data System grades 4-5, in the subgroup of nodules ≤1.5 cm, combined FNA/ CNB showed the best sensitivity in predicting malignancy (91.4%), significantly higher than that of CNB (81.0%, P<0.001) and FNA (57.8%, P<0.001). However, in the subgroup of nodules >1.5 cm, the difference between combined FNA/CNB and CNB was not significant (84.2% vs. 78.9%, P=0.500). Conclusion Regardless of nodule size, combined FNA/CNB tended to yield lower proportions of inconclusive results than CNB or FNA alone and exhibited higher performance in diagnosing malignancy. The combined FNA/CNB technique may be a more valuable diagnostic method for nodules ≤1.5 cm and nodules with a risk of malignancy than CNB and FNA alone.

This study explored the protective effect of astragaloside Ⅳ(AS-Ⅳ) on oxygen-glucose deprivation(OGD)-induced autophagic injury in PC12 cells and its underlying mechanism. An OGD-induced autophagic injury model in vitro was established in PC12 cells. The cells were divided into a normal group, an OGD group, low-, medium-, and high-dose AS-Ⅳ groups, and a positive drug dexmedetomidine(DEX) group. Cell viability was measured using the MTT assay. Transmission electron microscopy was used to observe autophagosomes and autolysosomes, and the MDC staining method was used to assess the fluorescence intensity of autophagosomes. Western blot was conducted to determine the relative expression levels of functional proteins LC3-Ⅱ/LC3-Ⅰ, Beclin1, p-Akt/Akt, p-mTOR/mTOR, and HIF-1α. Compared with the normal group, the OGD group exhibited a significant decrease in cell viability(P<0.01), an increase in autophagosomes(P<0.01), enhanced fluorescence intensity of autophagosomes(P<0.01), up-regulated Beclin1, LC3-Ⅱ/LC3-Ⅰ, and HIF-1α(P<0.05 or P<0.01), and down-regulated p-Akt/Akt and p-mTOR/mTOR(P<0.05 or P<0.01). Compared with the OGD group, the low-and medium-dose AS-Ⅳ groups and the DEX group showed a significant increase in cell viability(P<0.01), decreased autophagosomes(P<0.01), weakened fluorescence intensity of autophagosomes(P<0.01), down-regulated Beclin1, LC3-Ⅱ/LC3-Ⅰ, and HIF-1α(P<0.05 or P<0.01), and up-regulated p-Akt/Akt and p-mTOR/mTOR(P<0.01). AS-Ⅳ at low and medium doses exerted a protective effect against OGD-induced autophagic injury in PC12 cells by activating the Akt/mTOR pathway, subsequently influencing HIF-1α. The high-dose AS-Ⅳ group did not show a statistically significant difference compared with the OGD group. This study provides a certain target reference for the prevention and treatment of OGD-induced cellular autophagic injury by AS-Ⅳ and accumulates laboratory data for the secondary development of Astragali Radix and AS-Ⅳ.
Rats ; Animals ; PC12 Cells ; Proto-Oncogene Proteins c-akt/genetics* ; Glucose/therapeutic use* ; Oxygen/metabolism* ; Beclin-1/pharmacology* ; TOR Serine-Threonine Kinases/metabolism* ; Autophagy ; Apoptosis ; Reperfusion Injury/drug therapy*