Development and validation of a deep learning-based survival prediction model for pediatric glioma patients: A retrospective study using the SEER database and Chinese data.

OBJECTIVE: Develop a time-dependent deep learning model to accurately predict the prognosis of pediatric glioma patients, which can assist clinicians in making precise treatment decisions and reducing patient risk. STUDY DESIGN: The study involved pediatric glioma patients from the Surveillance, Epidemiology, and End Results (SEER) Registry (2000-2018) and Tangdu Hospital in China (2010-2018) within specific time frames. For training, we selected two neural network-based algorithms (DeepSurv, neural multi-task logistic regression [N-MTLR]) and one ensemble learning-based algorithm (random survival forest [RSF]). Additionally, a multivariable Cox proportional hazard (CoxPH) model was developed for comparison purposes. The SEER dataset was randomly divided into 80 % for training and 20 % for testing, while the Tangdu Hospital dataset served as an external validation cohort. Super-parameters were fine-tuned through 1000 repeated random searches and 5-fold cross-validation on the training cohort. Model performance was assessed using the concordance index (C-index), Brier score, and Integrated Brier Score (IBS). Furthermore, the accuracy of predicting survival at 1, 3, and 5 years was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and the area under the ROC curves (AUC). The generalization ability of the model was assessed using the C-index of the Tangdu Hospital data, ROC curves for 1, 3, and 5 years, and AUC values. Lastly, decision curve analysis (DCA) curves for 1, 3, and 5-year time frames are provided to assess the net benefits across different models. RESULTS: A total of 9532 patients with pediatric glioma were included in this study, comprising 9274 patients from the SEER database and 258 patients from Tangdu Hospital in China. The average age at diagnosis was 9.4 ± 6.2 years, and the average survival time was 96 ± 66 months. Through comprehensive performance comparison, the DeepSurv model demonstrated the highest effectiveness, with a C-index of 0.881 on the training cohort. Furthermore, it exhibited excellent accuracy in predicting the 1-year, 3-year, and 5-year survival rates (AUC: 0.903-0.939). Notably, the DeepSurv model also achieved remarkable performance and accuracy on the Chinese dataset (C-index: 0.782, AUC: 0.761-0.852). Comprehensive analysis of DeepSurv, N-MTLR, and RSF revealed that tumor stage, radiotherapy, histological type, tumor size, chemotherapy, age, and surgical method are all significant factors influencing the prognosis of pediatric glioma. Finally, an online version of the pediatric glioma survival predictor based on the DeepSurv model has been established and can be accessed through https://pediatricglioma-tangdu.streamlit.app. CONCLUSIONS: The DeepSurv model exhibits exceptional efficacy in predicting the survival of pediatric glioma patients, demonstrating strong performance in discrimination, calibration, stability, and generalization. By utilizing the online version of the pediatric glioma survival predictor, which is based on the DeepSurv model, clinicians can accurately predict patient survival and offer personalized treatment options.

TCGA Database-Based Screening of Tumor Microenvironment Immunomodulators Related to Bladder Cancer Prognosis.

OBJECTIVE: Bladder cancer (BC), as the most common malignant tumor of the urinary tract, has a complex biological behavior. Currently, there are still some limitations in the diagnosis and treatment of BC. Despite the great progress made in immunotherapy, there is still a lack of key genes for the diagnosis of BC. Therefore, it is particularly important to explore the differentially expressed genes (DEGs) and their effectiveness on prognosis of BC with different tumor microenvironment scores. METHODS: The gene expression dataset of BC was downloaded from the Cancer Genome Atlas (TCGA) database. The correlation between clinicopathological characteristics of patients and scores of immune and stromal components was analyzed. Patients were divided into high and low score groups according to their tumor microenvironment score (Immune score, Stromal score, ESTIMATE score). DEGs between high and low score groups were identified using R software and then subjected to enrichment analyses to assess their potential biological functions and signaling pathways. The protein-protein interaction (PPI) network was constructed using the STRING database to further identify hub genes. The expression levels of hub genes in BC were verified by TCGA database. Subsequently, the hub genes were evaluated for overall survival (OS), disease-free survival (DFS), progression-free survival (PFS), and disease-specific survival (DSS), and corresponding forest plots were created. RESULTS: A total of 2346 DEGs were obtained, including 1120 up-regulated genes and 1226 down-regulated genes. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses found DEGs were mainly enriched in cell migration and immune-related pathways. Meanwhile, The PPI network finally yielded top 10 hub genes with predictive value, which included actin beta (ACTB), interleukin 6 (IL-6), Jun proto-oncogene (JUN), CD4 molecule (CD4), heat shock protein 90 alpha family class A member 1 (HSP90AA1), protein tyrosine phosphatase receptor type C (PTPRC), tumor protein p53 (TP53), SRC proto-oncogene (SRC), fibronectin 1 (FN1), and tumor necrosis factor (TNF). Among them, CD4, PTPRC, and SRC were potential protective factors for BC. CONCLUSION: The top 10 hub genes (ACTB, IL-6, JUN, CD4, HSP90AA1, PTPRC, TP53, SRC, FN1, TNF) obtained based on tumor microenvironment scores all had potential predictive value. Elevated expression of protective factors (CD4, PTPRC, and SRC) indicates better survival outcome of BC subjects. Further exploration of the molecular developmental mechanisms of these hub genes will help to develop novel personalized therapies and improve BC prognosis.

Effects of gut microbiome on type 1 diabetes susceptibility and complications: A large-scale bidirectional Mendelian randomization and external validation study.

AIM: To assess and verify the effect of the gut microbiome on the susceptibility and complications of type 1 diabetes (T1D). MATERIALS AND METHODS: To achieve this aim, a two-sample and reverse Mendelian randomization (MR) analysis was conducted. In addition, an external validation study was performed using individual microbiome data of patients with T1D from the gutMEGA datasets and the National Clinical Research Center for Metabolic Diseases. The circulating metabolites facilitated two-sample MR analysis, mediation and multivariable MR analysis to evaluate the direct relationship between the gut microbiome and T1D complications. RESULTS: The MR analysis results from the discovery and validation phases confirmed that Veillonellaceae can potentially reduce the susceptibility of T1D. In the gutMEGA dataset, the average relative abundance of Veillonellaceae in patients with T1D was 0.66%, compared with 1.09% in the controls. Furthermore, the external validation, which included 60 patients with T1D and 30 matched healthy controls, found that the median relative abundance of Veillonellaceae was also lower than controls at 1.10% (95% CI 0.50%-1.80%). Specifically, the Eubacterium coprostanoligenes group, known for its ability to regulate cholesterol, was significantly associated with a lower risk of developing renal, neurological and ophthalmic complications in T1D. Moreover, high cholesterol in small high-density lipoprotein and cholesteryl esters in high-density lipoprotein were associated with a reduced risk of T1D renal and ophthalmic complications. The mediation and multivariable MR analysis combining cholesterol indicated that the E. coprostanoligenes group is the most dominant factor influencing T1D complications. CONCLUSIONS: Our findings supported the potential causal effect of gut microbiota on the susceptibility and complications of T1D.

Effect of Myelin Debris on the Phenotypic Transformation of Astrocytes after Spinal Cord Injury in Rats.

After spinal cord injury (SCI), the accumulation of myelin debris can serve as proinflammatory agents, hindering axon regrowth and exacerbating damage. While astrocytes have been implicated in the phagocytosis of myelin debris, the impact of this process on the phenotypic transformation of astrocytes and their characteristics following SCI in rats is not well understood. Here, we demonstrated that the conditioned medium of myelin debris can trigger apoptosis in rat primary astrocytes in vitro. Using a compressional SCI model in rats, we observed that astrocytes can engulf myelin debris through ATP-binding cassette transporter sub-family A member 1 (ABCA1), and these engulfed cells tend to transform into A1 astrocytes, as indicated by C3 expression. At 4 days post-injury (dpi), astrocytes rapidly transitioned into A1 astrocytes and maintained this phenotype from 4 to 28 dpi, while A2 astrocytes, characterized by S100, were only detected at 14 and 28 dpi. Reactive astrocytes, identified by Nestin, emerged at 4 and 7 dpi, whereas scar-forming astrocytes, marked by N-cadherin, were evident at 14 and 28 dpi. This study illustrates the distribution patterns of astrocyte subtypes and the potential interplay between astrocytes and myelin debris after SCI in rats. We emphasize that myelin debris can induce astrocyte apoptosis in vitro and promote the transformation of astrocytes into A1 astrocytes in vivo. These two classification methods are not mutually exclusive, but rather complementary.

Estimating the effect of lipid-lowering agents on novel subtypes of adult-onset diabetes.

AIMS: The aims of the present study were to assess the effects of lipid-lowering drugs [HMG-CoA reductase inhibitors, proprotein convertase subtilisin/kexin type 9 inhibitors, and Niemann-Pick C1-Like 1 (NPC1L1) inhibitors] on novel subtypes of adult-onset diabetes through a Mendelian randomisation study. MATERIALS AND METHODS: We first inferred causal associations between lipid-related traits [including high-density lipoprotein cholesterol, low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), apolipoproteins A-I, and apolipoproteins B] and novel subtypes of adult-onset diabetes. The expression quantitative trait loci of drug target genes for three classes of lipid-lowering drugs, as well as genetic variants within or nearby drug target genes associated with LDL-C, were then utilised as proxies for the exposure of lipid-lowering drugs. Mendelian randomisation analysis was performed using summary data from genome-wide association studies of LDL-C, severe autoimmune diabetes, severe insulin-deficient diabetes (SIDD), severe insulin-resistant diabetes (SIRD), mild obesity-related diabetes (MOD), and mild age-related diabetes. RESULTS: There was an association between HMGCR-mediated LDL-C and the risk of SIRD [odds ratio (OR) = 0.305, 95% confidence interval (CI) = 0.129-0.723; p = 0.007], and there was an association of PCSK9-mediated LDL-C with the risk of SIDD (OR = 0.253, 95% CI = 0.120-0.532; p < 0.001) and MOD (OR = 0.345, 95% CI = 0.171-0.696; p = 0.003). Moreover, NPC1L1-mediated LDL-C (OR = 0.109, 95% CI = 0.019-0.613; p = 0.012) and the increased expression of NPC1L1 gene in blood (OR = 0.727, 95% CI = 0.541-0.977; p = 0.034) both showed a significant association with SIRD. These results were further confirmed by sensitivity analyses. CONCLUSIONS: In summary, the different lipid-lowering medications have a specific effect on the increased risk of different novel subtypes of adult-onset diabetes.

Multi-Path Fusion in SFCF-Net for Enhanced Multi-Frequency Electrical Impedance Tomography.

Multi-frequency electrical impedance tomography (mfEIT) offers a nondestructive imaging technology that reconstructs the distribution of electrical characteristics within a subject based on the impedance spectral differences among biological tissues. However, the technology faces challenges in imaging multi-class lesion targets when the conductivity of background tissues is frequency-dependent. To address these issues, we propose a spatial-frequency cross-fusion network (SFCF-Net) imaging algorithm, built on a multi-path fusion structure. This algorithm uses multi-path structures and hyper-dense connections to capture both spatial and frequency correlations between multi-frequency conductivity images, which achieves differential imaging for lesion targets of multiple categories through cross-fusion of information. According to both simulation and physical experiment results, the proposed SFCF-Net algorithm shows an excellent performance in terms of lesion imaging and category discrimination compared to the weighted frequency-difference, U-Net, and MMV-Net algorithms. The proposed algorithm enhances the ability of mfEIT to simultaneously obtain both structural and spectral information from the tissue being examined and improves the accuracy and reliability of mfEIT, opening new avenues for its application in clinical diagnostics and treatment monitoring.

hBcl2 overexpression in BMSCs enhances resistance to myelin debris-induced apoptosis and facilitates neuroprotection after spinal cord injury in rats.

After spinal cord injury (SCI), the accumulation of myelin debris at the lesion exacerbates cell death and hinders axonal regeneration. Transplanted bone marrow mesenchymal stem cells (BMSCs) have been proven to be beneficial for SCI repair, but they are susceptible to apoptosis. It remains unclear whether this apoptotic process is influenced by myelin debris. Here, we constructed rat BMSCs overexpressing human B-cell lymphoma 2 (hBcl2) alone (hBcl2 group), BMSCs overexpressing hBcl2 with an endoplasmic reticulum-anchored segment (hBcl2-cb) (cb group), and a negative control group (NC group) for transplantation in this study. Immunocytochemistry staining validated the successful expression of hBcl2 in BMSCs within the hBcl2 group and cb group. All BMSCs from each group exhibited the ability to phagocytize myelin debris. Nevertheless, only BMSCs derived from the hBcl2 group exhibited heightened resistance to apoptosis and maintained prolonged viability for up to 5 days when exposed to myelin debris. Notably, overexpression of hBcl2 protein, rather than its endoplasmic reticulum-anchored counterpart, significantly enhanced the resistance of BMSCs against myelin debris-induced apoptosis. This process appeared to be associated with the efficient degradation of myelin debris through the Lamp1+ lysosomal pathway in the hBcl2 group. In vivo, the hBcl2 group exhibited significantly higher numbers of surviving cells and fewer apoptotic BMSCs compared to the cb and NC groups following transplantation. Furthermore, the hBcl2 group displayed reduced GFAP+ glial scarring and greater preservation of NF200+ axons in the lesions of SCI rats. Our results suggest that myelin debris triggers apoptosis in transplanted BMSCs, potentially elucidating the low survival rate of these cells after SCI. Consequently, the survival rate of transplanted BMSCs is improved by hBcl2 overexpression, leading to enhanced preservation of axons within the injured spinal cord.

Early detection of acute ischemic stroke using Contrast-enhanced electrical impedance tomography perfusion.

A cerebral contrast-enhanced electrical impedance tomography perfusion method is developed for acute ischemic stroke during intravenous thrombolytic therapy. Several clinical contrast agents with stable impedance characteristics and high-conductivity contrast were screened experimentally as electrical impedance contrast agent candidates. The electrical impedance tomography perfusion method was tested on rabbits with focal cerebral infarction, and its capability for early detection was verified based on perfusion images. The experimental results showed that ioversol 350 performed significantly better as an electrical impedance contrast agent than other contrast agents (p < 0.01). Additionally, perfusion images of focal cerebral infarction in rabbits confirmed that the electrical impedance tomography perfusion method could accurately detect the location and area of different cerebral infarction lesions (p < 0.001). Therefore, the cerebral contrast-enhanced electrical impedance tomography perfusion method proposed herein combines traditional, dynamic continuous imaging with rapid detection and could be applied as an early, rapid-detection, auxiliary, bedside imaging method for patients after a suspected ischemic stroke in both prehospital and in-hospital settings.

Effective Electrical Impedance Tomography Based on Enhanced Encoder-Decoder Using Atrous Spatial Pyramid Pooling Module.

Electrical impedance tomography (EIT) is a noninvasive and radiation-free imaging method. As a "soft-field" imaging technique, in EIT, the target signal in the center of the measured field is frequently swamped by the target signal at the edge, which restricts its further application. To alleviate this problem, this study presents an enhanced encoder-decoder (EED) method with an atrous spatial pyramid pooling (ASPP) module. The proposed method enhances the ability to detect central weak targets by constructing an ASPP module that integrates multiscale information in the encoder. The multilevel semantic features are fused in the decoder to improve the boundary reconstruction accuracy of the center target. The average absolute error of the imaging results by the EED method reduced by 82.0%, 83.6%, and 36.5% in simulation experiments and 83.0%, 83.2%, and 36.1% in physical experiments compared with the errors of the damped least-squares algorithm, Kalman filtering method, and U-Net-based imaging method, respectively. The average structural similarity improved by 37.3%, 42.9%, and 3.6%, and 39.2%, 45.2%, and 3.8% in the simulation and physical experiments, respectively. The proposed method provides a practical and reliable means of extending the application of EIT by solving the problem of weak central target reconstruction under the effect of strong edge targets in EIT.