ABSTRACT
Congenital heart disease (CHD) is a common birth defect in children. Intelligent auscultation algorithms have been proven to reduce the subjectivity of diagnoses and alleviate the workload of doctors. However, the development of this algorithm has been limited by the lack of reliable, standardized, and publicly available pediatric heart sound databases. Therefore, the objective of this research is to develop a large-scale, high-standard, high-quality, and accurately labeled pediatric CHD heart sound database. METHOD: From 2020 to 2022, we collaborated with experienced cardiac surgeons from three general children's hospitals to collect heart sound signals from 1259 participants using electronic stethoscopes. To ensure the accuracy of the labels, the labels for all data were confirmed by two cardiac experts. To establish the baseline of ZCHsound, we extracted 84 features and used machine learning models to evaluate the performance of the classification task. RESULTS: The ZCHSound database was divided into two datasets: one is a high-quality, filtered clean heart sound dataset, and the other is a low-quality, noisy heart sound dataset. In the evaluation of the high-quality dataset, our random forest ensemble model achieved an F1 score of 90.3% in the classification task of normal and pathological heart sounds. CONCLUSION: This study has successfully established a large-scale, high-quality, rigorously standardized pediatric CHD sound database with precise disease diagnosis. This database not only provides important learning resources for clinical doctors in auscultation knowledge but also offers valuable data support for algorithm engineers in developing intelligent auscultation algorithms.
Subject(s)
Databases, Factual , Heart Defects, Congenital , Heart Sounds , Signal Processing, Computer-Assisted , Humans , Heart Defects, Congenital/physiopathology , Heart Defects, Congenital/diagnostic imaging , Heart Sounds/physiology , Child , Child, Preschool , Infant , Algorithms , Male , Machine Learning , Female , Infant, Newborn , AdolescentABSTRACT
PURPOSE: To explore the predictive value of radiomics in predicting stroke-associated pneumonia (SAP) in acute ischemic stroke (AIS) patients and construct a prediction model based on clinical features and DWI-MRI radiomics features. METHODS: Univariate and multivariate logistic regression analyses were used to identify the independent clinical predictors for SAP. Pearson correlation analysis and the least absolute shrinkage and selection operator with ten-fold cross-validation were used to calculate the radiomics score for each feature and identify the predictive radiomics features for SAP. Multivariate logistic regression was used to combine the predictive radiomics features with the independent clinical predictors. The prediction performance of the SAP models was evaluated using receiver operating characteristics (ROC), calibration curves, decision curve analysis, and subgroup analyses. RESULTS: Triglycerides, the neutrophil-to-lymphocyte ratio, dysphagia, the National Institutes of Health Stroke Scale (NIHSS) score, and internal carotid artery stenosis were identified as clinically independent risk factors for SAP. The radiomics scores in patients with SAP were generally higher than in patients without SAP (P < 0. 05). There was a linear positive correlation between radiomics scores and NIHSS scores, as well as between radiomics scores and infarct volume. Infarct volume showed moderate performance in predicting the occurrence of SAP, with an AUC of 0.635. When compared with the other models, the combined prediction model achieved the best area under the ROC (AUC) in both training (AUC = 0.859, 95% CI 0.759-0.936) and validation (AUC = 0.830, 95% CI 0.758-0.896) cohorts (P < 0.05). The calibration curves and decision curve analysis further confirmed the clinical value of the nomogram. Subgroup analysis showed that this nomogram had potential generalization ability. CONCLUSION: The addition of the radiomics features to the clinical model improved the prediction of SAP in AIS patients, which verified its feasibility.
Subject(s)
Ischemic Stroke , Pneumonia , Stroke , United States , Humans , Feasibility Studies , Radiomics , Stroke/complications , Stroke/diagnostic imaging , Stroke/epidemiology , InfarctionABSTRACT
INTRODUCTION: Stroke is a leading cause of mortality and disability worldwide. Recurrent strokes result in prolonged hospitalisation and worsened functional outcomes compared with the initial stroke. Thus, it is critical to identify patients who are at high risk of stroke recurrence. This study is positioned to develop and validate a prediction model using radiomics data and machine learning methods to identify the risk of stroke recurrence in patients with acute ischaemic stroke (AIS). METHODS AND ANALYSIS: A total of 1957 patients with AIS will be needed. Enrolment at participating hospitals will continue until the required sample size is reached, and we will recruit as many participants as possible. Multiple indicators including basic clinical data, image data, laboratory data, CYP2C19 genotype and follow-up data will be assessed at various time points during the registry, including baseline, 24 hours, 7 days, 1 month, 3 months, 6 months, 9 months and 12 months. The primary outcome was stroke recurrence. The secondary outcomes were death events, prognosis scores and adverse events. Imaging images were processed using deep learning algorithms to construct a programme capable of automatically labelling the lesion area and extracting radiomics features. The machine learning algorithms will be applied to integrate cross-scale, multidimensional data for exploratory analysis. Then, an ischaemic stroke recurrence prediction model of the best performance for patients with AIS will be established. Calibration, receiver operating characteristic and decision curve analyses will be evaluated. ETHICS AND DISSEMINATION: This study has received ethical approval from the Medical Ethics Committee of the Second Affiliated Hospital of Nanchang University (medical research review No.34/2021), and informed consent will be obtained voluntarily. The research findings will be disseminated through publication in journals and presented at conferences. TRIAL REGISTRATION NUMBER: ChiCTR2200055209.
Subject(s)
Brain Ischemia , Ischemic Stroke , Stroke , Humans , Stroke/complications , Brain Ischemia/complications , Prospective Studies , Ischemic Stroke/diagnostic imaging , Ischemic Stroke/complications , Machine Learning , Observational Studies as Topic , Multicenter Studies as TopicABSTRACT
Purpose: To investigate the predictive value of various inflammatory biomarkers in patients with acute ischemic stroke (AIS) and evaluate the relationship between stroke-associated pneumonia (SAP) and the best predictive index. Patients and Methods: We calculated the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), prognostic nutritional index (PNI), systemic inflammation response index (SIRI), systemic immune inflammation index (SII), Glasgow prognostic score (GPS), modified Glasgow prognostic score (mGPS), and prognostic index (PI). Variables were selectively included in the logistic regression analysis to explore the associations of NLR, PLR, MLR, PNI, SIRI, SII, GPS, mGPS, and PI with SAP. We assessed the predictive performance of biomarkers by analyzing receiver operating characteristic (ROC) curves. We further used restricted cubic splines (RCS) to investigate the association. Next, we conducted subgroup analyses to investigate whether specific populations were more susceptible to NLR. Results: NLR, PLR, MLR, SIRI, SII, GPS, mGPS, and PI increased significantly in SAP patients, and PNI was significantly decreased. After adjustment for potential confounders, the association of inflammatory biomarkers with SAP persisted. NLR showed the most favorable discriminative performance and was an independent risk factor predicting SAP. The RCS showed an increasing nonlinear trend of SAP risk with increasing NLR. The AUC of the combined indicator of NLR and C-reactive protein (CRP) was significantly higher than those of NLR and CRP alone (DeLong test, P<0.001). Subgroup analyses suggested good generalizability of the predictive effect. Conclusion: NLR, PLR, MLR, PNI, SIRI, SII, GPS, mGPS, and PI can predict the occurrence of SAP. Among the indices, the NLR was the best predictor of SAP occurrence. It can therefore be used for the early identification of SAP.
Subject(s)
Ischemic Stroke , Pneumonia , Stroke , Humans , Stroke/complications , Pneumonia/complications , Biomarkers , Inflammation , C-Reactive ProteinABSTRACT
Background: With the aggregation of clinical data and the evolution of computational resources, artificial intelligence-based methods have become possible to facilitate clinical diagnosis. For congenital heart disease (CHD) detection, recent deep learning-based methods tend to achieve classification with few views or even a single view. Due to the complexity of CHD, the input images for the deep learning model should cover as many anatomical structures of the heart as possible to enhance the accuracy and robustness of the algorithm. In this paper, we first propose a deep learning method based on seven views for CHD classification and then validate it with clinical data, the results of which show the competitiveness of our approach. Methods: A total of 1411 children admitted to the Children's Hospital of Zhejiang University School of Medicine were selected, and their echocardiographic videos were obtained. Then, seven standard views were selected from each video, which were used as the input to the deep learning model to obtain the final result after training, validation and testing. Results: In the test set, when a reasonable type of image was input, the area under the curve (AUC) value could reach 0.91, and the accuracy could reach 92.3%. During the experiment, shear transformation was used as interference to test the infection resistance of our method. As long as appropriate data were input, the above experimental results would not fluctuate obviously even if artificial interference was applied. Conclusions: These results indicate that the deep learning model based on the seven standard echocardiographic views can effectively detect CHD in children, and this approach has considerable value in practical application.
ABSTRACT
AIMS: The aim of this study was to determine whether the testing strategy for clopidogrel and/or aspirin resistance using CYP2C19 genotyping or urinary 11-dhTxB2 testing has an impact on clinical outcomes. METHODS: A multicentre, randomized, controlled trial was conducted at 14 centres in China from 2019 to 2021. For the intervention group, a specific antiplatelet strategy was assigned based on the CYP2C19 genotype and 11-dhTxB2, a urinary metabolite of aspirin, and the control group received nonguided (ie, standard of care) treatment. 11-dhTXB2 is a thromboxane A2 metabolite that can help quantify the effects of resistance to aspirin in individuals after ingestion. The primary efficacy outcome was new stroke, the secondary efficacy outcome was a poor functional prognosis (a modified Rankin scale score ≥3), and the primary safety outcome was bleeding, all within the 90-day follow-up period. RESULTS: A total of 2815 patients were screened and 2663 patients were enrolled in the trial, with 1344 subjects assigned to the intervention group and 1319 subjects assigned to the control group. A total of 60.1% were carriers of the CYP2C19 loss-of-function allele (*2, *3) and 8.71% tested positive for urinary 11-dhTxB2- indicating aspirin resistance in the intervention group. The primary outcome was not different between the intervention and control groups (P = .842). A total of 200 patients (14.88%) in the intervention group and 240 patients (18.20%) in the control group had a poor functional prognosis (hazard ratio 0.77, 95% confidence interval [CI] 0.63 to 0.95, P = .012). Bleeding events occurred in 49 patients (3.65%) in the intervention group and 72 patients (5.46%) in the control group (hazard ratio 0.66, 95% CI 0.45 to 0.95, P = .025). CONCLUSIONS: Personalized antiplatelet therapy based on the CYP2C19 genotype and 11-dhTxB2 levels was associated with favourable neurological function and reduced bleeding risk in acute ischaemic stroke and transient ischaemic attack patients. The results may help support the role of CYP2C19 genotyping and urinary 11-dhTxB2 testing in the provision of precise clinical treatment.
ABSTRACT
Purpose: This study aimed to investigate the value of a machine learning-based magnetic resonance imaging (MRI) radiomics model in predicting the risk of recurrence within 1 year following an acute ischemic stroke (AIS). Methods: The MRI and clinical data of 612 patients diagnosed with AIS at the Second Affiliated Hospital of Nanchang University from March 1, 2019, to March 5, 2021, were obtained. The patients were divided into recurrence and non-recurrence groups according to whether they had a recurrent stroke within 1 year after discharge. Randomized splitting was used to divide the data into training and validation sets using a ratio of 7:3. Two radiologists used the 3D-slicer software to label the lesions on brain diffusion-weighted (DWI) MRI sequences. Radiomics features were extracted from the annotated images using the pyradiomics software package, and the features were filtered using the Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis. Four machine learning algorithms, logistic regression (LR), Support Vector Classification (SVC), LightGBM, and Random forest (RF), were used to construct a recurrence prediction model. For each algorithm, three models were constructed based on the MRI radiomics features, clinical features, and combined MRI radiomics and clinical features. The sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC) were used to compare the predictive efficacy of the models. Results: Twenty features were selected from 1,037 radiomics features extracted from DWI images. The LightGBM model based on data with three different features achieved the best prediction accuracy from all 4 models in the validation set. The LightGBM model based solely on radiomics features achieved a sensitivity, specificity, and AUC of 0.65, 0.671, and 0.647, respectively, and the model based on clinical data achieved a sensitivity, specificity, and AUC of 0.7, 0.799, 0.735, respectively. The sensitivity, specificity, and AUC of the LightGBM model base on both radiomics and clinical features achieved the best performance with a sensitivity, specificity, and AUC of 0.85, 0.805, 0.789, respectively. Conclusion: The ischemic stroke recurrence prediction model based on LightGBM achieved the best prediction of recurrence within 1 year following an AIS. The combination of MRI radiomics features and clinical data improved the prediction performance of the model.
Subject(s)
Cholecystectomy, Laparoscopic , Laparoscopy , Anesthesia, Local , Anesthetics, Local , Humans , Pain, PostoperativeABSTRACT
Transient electronics is an emerging class of electronic devices that can physically degrade or disintegrate after a stable period of service, showing a vast prospect in applications of "green" consumer electronics, hardware-secure devices, medical implants, etc. Complementary metal-oxide-semiconductor (CMOS) technology is dominant in integrated circuit design for its advantages of low static power consumption, high noise immunity, and simple design layout, which also work and are highly preferred for transient electronics. However, the performance of complementary transient electronics is severely restricted by the confined selection of transient materials and compatible fabrication strategies. Here, we report the realization of high-performance transient complementary electronics based on carbon nanotube thin films via a reliable electrostatic doping method. Under a low operating voltage of 2 V, on a 1.5 µm-thick water-soluble substrate made of poly(vinyl alcohol), the width-normalized on-state currents of the p-type and n-type transient thin-film transistors (TFTs) reach 4.5 and 4.7 µA/µm, and the width-normalized transconductances reach 2.8 and 3.7 µS/µm, respectively. Meanwhile, these TFTs show small subthreshold swings no more than 108 mV/dec and current on/off ratios above 106 with good uniformity. Transient CMOS inverters, as basic circuit components, are demonstrated with a voltage gain of 24 and a high noise immunity of 67.4%. Finally, both the degradation of the active components and the disintegration of the functional system are continuously monitored with nontraceable remains after 10 and 5 h, respectively.
ABSTRACT
Flexible sensors have attracted increasing research interest due to their broad application potential in the fields of human-computer interaction, medical care, sports monitoring, etc. Constructing an integrated sensor system with high performance and being capable of discriminating different stimuli remains a challenge. Here, we proposed a flexible integrated sensor system for motion monitoring that can measure bending strain and pressure independently with a low-cost and simple fabrication process. The resistive bending strain sensor in the system is fabricated by sintering polyimide (PI), demonstrating a gauge factor of 9.54 and good mechanical stability, while the resistive pressure sensor is constructed based on a composite structure of silver nanowires (AgNWs) and polydimethylsiloxane (PDMS)-expandable microspheres with a tunable sensitivity and working range. Action recognition is demonstrated by attaching the flexible integrated sensor system on the wrist with independent strain and pressure information recorded from corresponding sensors. It shows a great application potential in motion monitoring and intelligent human-machine interfaces.
Subject(s)
Nanowires , Wearable Electronic Devices , Humans , Motion , Silver , WristABSTRACT
BACKGROUND: Clopidogrel and aspirin are key intervention for acute ischemic stroke (AIS) and transient ischemic attack (TIA). However, with increased clinical application, many patients have shown clopidogrel resistance (CR) and/or aspirin resistance (AR) that affect antiplatelet therapy on AIS/TIA. At present, there is no research reported on personalized antiplatelet therapy guidelines for patients with CR and/or AR. Our study aims to assess the effect of personalized antiplatelet therapy based on CYP2C19 genotype and urine 11-dhTxB2 tests in patients with AIS or TIA. METHODS: This is a multi-center randomized controlled trial. Eligible patients with AIS/TIA from 14 comprehensive hospitals in Jiangxi province will be recruited after obtaining informed consent. Participants will be randomly divided into the intervention group and the control group at a ratio of 1:1. personalized antiplatelet therapy based on the CYP2C19 genotype/urine11-dhTxB2 tests will be given to the intervention group. Demographics, disease history, laboratory investigations, therapys, physiological tests, imaging reports and other clinical features will be collected. Clinical outcomes including stroke recurrence, Modified Rankin Scale (mRS) score, bleeding events and all-cause mortality will be assessed at the 1st, 3rd, 6th, and 12th-month post-discharge. DISCUSSION: Our study will conduct free antiplatelet resistance tests and personalized antiplatelet therapy for AIS/TIA patients with CR/AR, ultimately evaluating personalized therapy effectiveness through one-year follow-up. The research results will help to assess the impact of personalized antiplatelet therapy on the prognosis of stroke, thus providing reference for precise clinical treatment.
Subject(s)
Brain Ischemia , Ischemic Attack, Transient , Ischemic Stroke , Stroke , Aftercare , Aspirin/therapeutic use , Brain Ischemia/drug therapy , Clopidogrel/therapeutic use , Humans , Ischemic Attack, Transient/drug therapy , Multicenter Studies as Topic , Patient Discharge , Platelet Aggregation Inhibitors/therapeutic use , Randomized Controlled Trials as Topic , Stroke/drug therapyABSTRACT
Steep canyons surrounded by high mountains resulting from large-scale landslides characterize the study area located in the southeastern part of the Tibetan Plateau. A total of 1766 large landslides were identified based on integrated remote sensing interpretations utilizing multisource satellite images and topographic data that were dominated by 3 major regional categories, namely, rockslides, rock falls, and flow-like landslides. The geographical detector method was applied to quantitatively unveil the spatial association between the landslides and 12 environmental factors through computation of the q values based on spatially stratified heterogeneity. Meanwhile, a certainty factor (CF) model was used for comparison. The results indicate that the q values of the 12 influencing factors vary obviously, and the dominant factors are also different for the 3 types of landslides, with annual mean precipitation (AMP) being the dominant factor for rockslide distribution, elevation being the dominant factor for rock fall distribution and lithology being the dominant factor for flow-like distribution. Integrating the results of the factor detector and ecological detector, the AMP, annual mean temperature (AMT), elevation, river density, fault distance and lithology have a stronger influence on the spatial distribution of landslides than other factors. Furthermore, the factor interactions can significantly enhance their interpretability of landslides, and the top 3 dominant interactions were revealed. Based on statistics of landslide discrepancies with respect to diverse stratification of each factor, the high-risk zones were identified for 3 types of landslides, and the results were contrasted with the CF model. In conclusion, our method provides an objective framework for landslide prevention and mitigation through quantitative, spatial and statistical analyses in regions with complex terrain.
Subject(s)
Geographic Information Systems , Rivers , TibetABSTRACT
In this paper, novel multiaction antibacterial nanofibrous membranes containing apatite, Ag, AgBr and TiO2 as four active components were fabricated by an electrospinning technique. In this antibacterial membrane, each component serves a different function: the hydroxyapatite acts as the adsorption material for capturing bacteria, the Ag nanoparticles act as the release-active antibacterial agent, the AgBr nanoparticles act as the visible sensitive and release-active antibacterial agent, and the TiO2 acts as the UV sensitive antibacterial material and substrate for other functional components. Using E. coli as the typical testing organism, such multicomponent membranes exhibit excellent antimicrobial activity under UV light, visible light or in a dark environment. The significant antibacterial properties may be due to the synergetic action of the four major functional components, and the unique porous structure and high surface area of the nanofibrous membrane. It takes only 20 min for the bacteria to be completely (99.9%) destroyed under visible light. Even in a dark environment, about 50 min is enough to kill all of the bacteria. Compared to the four component system in powder form reported previously, the addition of the electrospun membrane could significantly improve the antibacterial inactivation of E. coli under the same evaluation conditions. Besides the superior antimicrobial capability, the permanence of the antibacterial activity of the prepared free-standing membranes was also demonstrated in repeated applications.