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

Artemisia argyi (A. argyi), a plant with a longstanding history as a raw material for traditional medicine and functional diets in Asia, has been used traditionally to bathe and soak feet for its disinfectant and itch-relieving properties. Despite its widespread use, scientific evidence validating the antifungal efficacy of A. argyi water extract (AAWE) against dermatophytes, particularly Trichophyton rubrum, Trichophyton mentagrophytes, and Microsporum gypseum, remains limited. This study aimed to substantiate the scientific basis of the folkloric use of A. argyi by evaluating the antifungal effects and the underlying molecular mechanisms of its active subfraction against dermatophytes. The results indicated that AAWE exhibited excellent antifungal effects against the three aforementioned dermatophyte species. The subfraction AAWE6, isolated using D101 macroporous resin, emerged as the most potent subfraction. The minimum inhibitory concentrations (MICs) of AAWE6 against T. rubrum, M. gypseum, and T. mentagrophytes were 312.5, 312.5, and 625 μg·mL^-1, respectively. Transmission electron microscopy (TEM) results and assays of enzymes linked to cell wall integrity and cell membrane function indicated that AAWE6 could penetrate the external protective barrier of T. rubrum, creating breaches ("small holes"), and disrupt the internal mitochondrial structure ("granary"). Furthermore, transcriptome data, quantitative real-time PCR (RT-qPCR), and biochemical assays corroborated the severe disruption of mitochondrial function, evidenced by inhibited tricarboxylic acid (TCA) cycle and energy metabolism. Additionally, chemical characterization and molecular docking analyses identified flavonoids, primarily eupatilin (131.16 ± 4.52 mg·g^-1) and jaceosidin (4.17 ± 0.18 mg·g^-1), as the active components of AAWE6. In conclusion, the subfraction AAWE6 from A. argyi exerts antifungal effects against dermatophytes by disrupting mitochondrial morphology and function. This research validates the traditional use of A. argyi and provides scientific support for its anti-dermatophytic applications, as recognized in the Chinese patent (No. ZL202111161301.9).
Antifungal Agents/chemistry* ; Arthrodermataceae ; Artemisia/chemistry* ; Molecular Docking Simulation ; Mitochondria ; Microbial Sensitivity Tests

Sj?gren′s syndrome is a chronic autoimmune inflammatory disease characterized by exocrine gland and extraglandular involvement. Cases of Sj?gren′s syndrome-associated aseptic meningitis (SS-AM) are relatively rare, and a case of recurrent aseptic meningitis with leukopenia and mild anemia associated with primary Sj?gren′s syndrome is reported, whose symptoms basically disappeared after treatment with prednison and hydroxychloroquine. The purpose of reporting this case is to raise awareness of SS-AM among fellow clinicians.

A three-dimensional(3D)medical image segmentation network(CA-SegResNet)which incorporates a 3D coordinate attention mechanism is proposed to address the issue of segmenting identified vertebral bones from spinal computed tomography(CT)images.The network extracts image features through a deep residual convolutional neural network and fuses the feature maps from each encoder layer with the input of the corresponding decoder layer.Subsequently,a 3D coordinate attention module is introduced to capture inter-channel relationships as well as directional and positional information,establishing long-range dependencies across different spatial directions,thereby enabling precise segmentation of the identified vertebral bones.For the segmentation tasks involving the identified cervical vertebra(the 7th cervical vertebra)and the identified thoracic vertebra(the 12th thoracic vertebra)across 105 cases,CA-SegResNet achieves average Dice similarity coefficients(DSC)of 0.934 5 and 0.918 9 on the test set,with average Hausdorff distances(HD)of 7 and 8 mm.Compared with U-Net results,the average DSC is improved by 0.014 5 and 0.0463,while average HD is reduced by 176 and 388 mm.The results demonstrate that the network can realize the precise segmentation of identified vertebral bones from CT images.

Objective:To analyze the clinical characteristics and prognosis of patients with stiff-person syndrome (SPS) associated with glutamic acid decarboxylase (GAD) antibodies.Methods:A retrospective analysis was conducted on demographic characteristics, clinical manifestations, auxiliary examination results, treatment, and prognosis of patients with GAD antibody-related SPS treated at Peking Union Medical College Hospital from January 2015 to July 2023.Results:A total of 33 patients were included, comprising 26 females (78.8%) and 7 males (21.2%), with an onset age of (42±12) years and a disease duration of 24.0 (10.5, 37.5) months. Two cases (6.1%) were diagnosed with tumors, including 1 case with invasive thymoma and 1 case with small cell lung cancer. The majority of patients (87.9%, 29/33) presented with stiffness of trunk and proximal limb muscles, 42.4% (14/33) of patients exhibited episodic spasm, and 54.5% (18/33) of patients were triggered by stimuli such as sound and light. Babinski or Chaddock reflexes were elicited in 33.3% (11/33) of patients. Some patients (36.4%, 16/33) had concurrent limbic encephalitis/epilepsy or cerebellar ataxia (referred to as complex SPS). The median cerebrospinal fluid (CSF) white blood cell count was 2×10 6/L [quartile: 1×10 6/L, 6×10 6/L; range: (0-30)×10 6/L], with mild elevation in 28.0% (7/25) of patients. Multi-channel surface electromyography in 14 out of 21 cases (66.7%) suggested synchronous contraction of agonist and antagonist muscles in a relaxed state. The modified Rankin Scale (mRS) score during the acute phase was 4 (3, 4). All patients received treatment with benzodiazepines or baclofen. Thirty patients (90.9%, 30/33) received first-line immunotherapy, 3 patients (9.1%, 3/33) received second-line immunotherapy with rituximab, and 14 (42.4%, 14/33) received mycophenolate mofetil as long-term immunotherapy. The follow-up period was 16 (10, 42) months, with a median best mRS score of 2; 66.7% (22/33) of patients had a favorable functional prognosis (mRS score≤2), and the recurrence rate was 30.0% (9/30). At the last follow-up, the median mRS score was 2, and 53.3% (16/30) of patients had a favorable functional prognosis. Prognosis was not significantly correlated with gender, age, clinical type, or CSF white blood cell level (all P>0.05). Conclusions:SPS is one of the main clinical phenotypes of GAD antibody-related neuroimmune diseases, commonly observed in middle-aged women, and exhibits a chronic progressive course. Only a minority of patients have concomitant tumors. The diagnosis relies on typical symptoms, GAD antibody testing, and electromyography examination. The initial immune therapy yields good results, but the prognosis for recurrent patients is poor.

Objective:To analyze the effectiveness and safety of the second course radiotherapy for unresectable colorectal cancer liver metastases.Methods:We retrospectively collected the data of 28 patients with unresectable colorectal cancer liver metastases who received the second course radiotherapy at Peking University Cancer Hospital and Institute from 2017 to 2023, to analyze the feasibility of re-irradiation.Results:For the 28 patients, the median follow-up time after re-irradiation was 20.2 months. The median time interval between the first- and second-course radiotherapy was 11.1 months. The median biologically effective doses of the first- and second-course radiotherapy were 100 Gy and 96 Gy, respectively. Stereotactic body radiotherapy was administered to 25 patients (89.3%) during the first course and 24 patients (85.7%) during the second course of radiotherapy. The mean equivalent dose in 2 Gy fractions to the normal liver was 10.1 Gy in the first-course radiotherapy and 7.9 Gy in the second-course radiotherapy. The complete response rate, partial response rate, and objective response rate after re-irradiation were 54.5%, 18.2%, and 72.7%, respectively. After re-irradiation, the 2-year cumulative local failure rate was 17.0% when calculated based on patients and 15.1% when calculated based on lesions, the 1-year progression-free survival rate was 27.4%, and the 3-year overall survival rate was 46.7%. The second-course radiotherapy was well tolerated, with most patients (75.0%) experiencing grade 1-2 acute adverse reactions and only one case (3.6%) experiencing grade 3 acute adverse events.Conclusions:Second course radiotherapy is an effective and safe treatment approach for selected patients with unresectable colorectal cancer liver metastases.

Paraneoplastic neurological syndromes (PNS) are heterogeneous disorders caused by autoimmune responses of cancer, which can affect any part of the nervous system. Anti-amphiphysin antibody is one of the high-risk PNS antibodies, which is usually associated with small cell lung cancer and breast cancer. However, extrapulmonary neuroendocrine carcinoma is rare in patients with anti-amphiphysin antibody. A case of anti-amphiphysin-associated paraneoplastic brainstem encephalitis with esophageal neuroendocrine carcinoma is reported. The tumor was detected by fluorine 18 fluorodeoxyglucose positron emission tomography and pathologically confirmed by gastroscopic biopsy. The patient′s neurological symptoms were partially improved after treatment of intravenous immunoglobulin and glucocorticoids. However, the disease prognosis is closely related to the accompanying tumor.

Objective:To explore the diagnostic value of 18F-deoxyglucose (FDG) PET/CT dual-time-point imaging (DTPI) in the diagnosis of aortic grafts infection (AGI). Methods:Forty-two patients with suspected AGI were prospectively recruited in this DTPI study from October 2014 to October 2021. There were 35(83%) males and 7 females, mean age (54±15) years old, range 22-79 years old. PET/CT image quality was scored as 5 grading scale. Semi-quantitative analysis of DTPI data was performed using maximum standardized uptake value (SUVmax) of suspected AGI lesions. The percentage of SUVmax change between initial and delayed images were recorded as retention index (RI). Management of Aortic Graft Infection Collaboration (MAGIC) criteria were used as the diagnostic reference criteria for AGI.Results:According to the MAGIC criteria, 27 patients (64%) were positive for AGI, and 15 patients (36%) were negative. The mean RI of AGI was higher than that of non-AGI ones[(26.7±18.9)% vs. (6.4 ±18.8)%, P<0.01]. The sensitivity, specificity, and accuracy of initial SUVmax ≥6 with the presence of AGI was 88.9%, 73.3%, and 83.3%, respectively. Delayed SUVmax ≥6 improved the sensitivity (96.3%) and accuracy (88.1%) for diagnosing AGI. DTPI with 15% increment as the optimal cut-off value of RI improved the specificity (93.3%) and accuracy (90.5%) for diagnosing AGI. Fifteen (56%, 15/27) AGI patients had improved image quality grading on the delayed images, leading to more accurately delineating the detailed extent of the infected aortic graft. Conclusion:18F-FDG PET/CT DTPI has better diagnostic performance for AGI than conventional Single-time-point PET/CT imaging by improving image quality as well as enhancing delineation of infected aortic graft extent.

ObjectiveThis study aimed to investigate the risk factors affecting the positive detection of Acinetobacter baumannii on the hands in medical staff of hospitals in Shanghai， and provide epidemiological evidence for the prevention and control of nosocomial infections. MethodsThe hands of doctors， nurses and care workers in key departments were sampled every quarter from 2018 to 2020 according toGB 15982‒2012 "Hospital Disinfection and Hygiene Standards". Separation and identification of A. baumannii were followed with sampling shortly. Information about the working years of sampling subjects， the hand sanitizers of which sampling subjects had used and the ingredients and actual using time of the hand sanitizers was collected while sampling. Finally， 709 samples were selected for this research after excluding the unqualified samples. ResultsThe positive detection of the hand samples was 7.05%. The logistic regression model suggested that the department， the time of using hand sanitizer， the hospital grade and occupational category were determinants of A. baumannii positive detection on hands in medical staff. The risk of A. baumannii positive detection in internal medicine department was 2.846 （95%CI：1.402‒5.776） times higher than that in intensive care unit while it was 3.357 （95%CI：1.349‒8.353） times higher in surgery department than that in intensive care unit. Regarding the use of hand sanitizer， the risk of A. baumannii positive detection was 3.076 （95%CI：1.534‒6.168） times higher in the staff used the hand sanitizer over 14 days than in the medical staff used the sanitizer within 14 days. The risk of A. baumannii positive detection in medical worker in secondary hospitals was 2.235（95%CI：1.088‒4.588）times than in tertiary hospitals. The risk of A. baumannii positive detection of care workers was 3.634 （95%CI：1.764‒7.484） times higher than nurses. ConclusionThe positive detection of hand samples was 7.05%. Department， the time of using hand sanitizer， the hospital grade and occupational category were determinants of A. baumannii positive detection on hands in medical staff. It was necessary to improve hand hygiene for medical staff， especially for care worker. Cleaning and disinfection need to be strengthened in internal department and surgery department.

Objective:To assess the long-term follow-up results of intensity-modulated radiotherapy (IMRT) combined with transcatheter arterial chemoembolization (TACE) and tyrosine kinase inhibitor (TKI) in patients with hepatocellular carcinoma (HCC) showing macrovascular invasion (MVI).Methods:A retrospective analysis was conducted for 63 patients with HCC showing MVI without distant metastasis treated in Peking University Cancer Hospital from October 2015 to October 2018. Among them 28 patients were treated with IMRT combined with TACE and sorafenib (Group A) and 35 patients were treated with IMRT combined with TACE (Group B). Propensity score matching (PSM) was applied to assess the progression-free survival (PFS) and the overall survival (OS) of both groups.Results:The median follow-up time was 62 months. Before PSM, the median OS of group A and B were 19.0 months and 15.2 months ( χ2=3.15, P=0.076), respectively, and the median PFS of groups A (10.7 months) was longer than that of group B (8.6 months; χ2=3.99, P=0.046). After PSM, the median OS of group A (30.6 months) was significantly longer than that of group B (15.2 months; χ2=5.34, P=0.023), and the PFS of groups A (12.5 months) was still longer than that of group B (8.3 months; χ2=4.79, P=0.026). In the whole group, 10 patients (15.9%) suffered from grade-3 hematologic toxicity, and seven patients (11.1%) experienced grade-3 hepatic toxicity. The incidence of skin reactions, hand-foot syndrome, and diarrhea in group A was higher than that in group B, but all these adverse events were grade 1-2. Moreover, no grade-4 adverse events, radiation-induced liver disease, and treatment-related mortality occurred in both groups. Conclusions:As demonstrated by the long-term follow-up result, IMRT combined with TACE and TKI could improve both the PFS and the OS of patients with HCC showing MVI after PSM.