RESUMEN
BACKGROUND: Catheter-related thrombosis (CRT) is a thrombotic complication associated with using central venous catheters (CVCs). Although risk factors for CRT were identified in children, no nomograms or predictive tools are available for the pediatric population with CVCs. This study aimed to develop and validate a prediction model of asymptomatic CRT in children with CVCs. METHODS: This retrospective observational study included consecutive pediatric patients who admitted to the Children's Hospital Zhejiang University School of Medicine and received CVCs between October and December 2021. RESULTS: This study included 669 patients, 553 (314 males, aged 22.00 [0.36, 180.00] months, 62 with CRT) were in the training set, and 116 (62 males, aged 15.00 [1.13, 156.00] months, 16 with CRT) were in the validation set. Multivariate logistic regression showed that a catheter time of 0-3 days (OR = 0.201, 95%CI: 0.081-0.497, P = 0.001), catheter time of 4-7 days (OR = 0.412, 95%CI: 0.176-0.964, P = 0.041), male (OR = 3.976, 95%CI: 1.864-4.483, P < 0.001), congenital heart diseases (OR = 0.277, 95%CI: 0.078-0.987, P = 0.048), postoperative (OR = 0.161, 95%CI: 0.072-0.360, P < 0.001), and femoral CVC (OR = 2.451, 95%CI: 1.129-5.318, P = 0.002) were independently associated with CRT. The nomogram incorporating these variables showed relatively good discrimination (AUC = 0.77, 95%CI: [0.65, 0.90]) and calibration abilities in the validation set, and the decision curve analysis (DCA) yielded a clinical net benefit. CONCLUSION: A prediction model for CRT in children with CVC was established based on catheter time, sex, diseases, postoperative, and catheter vein. The nomogram based on logistic regression model showed favorable predictive performance.
Asunto(s)
Catéteres Venosos Centrales , Nomogramas , Humanos , Masculino , Estudios Retrospectivos , Femenino , Niño , Adolescente , Lactante , Preescolar , Catéteres Venosos Centrales/efectos adversos , Factores de Riesgo , Trombosis/etiología , Trombosis/diagnóstico , Cateterismo Venoso Central/efectos adversos , Recién Nacido , Modelos LogísticosRESUMEN
Accurate teeth delineation on 3-D dental models is essential for individualized orthodontic treatment planning. Pioneering works like PointNet suggest a promising direction to conduct efficient and accurate 3-D dental model analyses in end-to-end learnable fashions. Recent studies further imply that multistream architectures to concurrently learn geometric representations from different inputs/views (e.g., coordinates and normals) are beneficial for segmenting teeth with varying conditions. However, such multistream networks typically adopt simple late-fusion strategies to combine features captured from raw inputs that encode complementary but fundamentally different geometric information, potentially hampering their accuracy in end-to-end semantic segmentation. This article presents a hierarchical cross-stream aggregation (HiCA) network to learn more discriminative point/cell-wise representations from multiview inputs for fine-grained 3-D semantic segmentation. Specifically, based upon our multistream backbone with input-tailored feature extractors, we first design a contextual cross-steam aggregation (CA) module conditioned on interstream consistency to boost each view's contextual representation learning jointly. Then, before the late fusion of different streams' outputs for segmentation, we further deploy a discriminative cross-stream aggregation (DA) module to concurrently update all views' discriminative representation learning by leveraging a specific graph attention strategy induced by multiview prototype learning. On both public and in-house datasets of real-patient dental models, our method significantly outperformed state-of-the-art (SOTA) deep learning methods for teeth semantic segmentation. In addition, extended experimental results suggest the applicability of HiCA to other general 3-D shape segmentation tasks. The code is available at https://github.com/ladderlab-xjtu/HiCA.
RESUMEN
The last-minute cancellation of surgeries profoundly affects patients and their families. This research aimed to forecast these cancellations using EMR data and meteorological conditions at the time of the appointment, using a machine learning approach. We retrospectively gathered medical data from 13 440 pediatric patients slated for surgery from 2018 to 2021. Following data preprocessing, we utilized random forests, logistic regression, linear support vector machines, gradient boosting trees, and extreme gradient boosting trees to predict these abrupt cancellations. The efficacy of these models was assessed through performance metrics. The analysis revealed that key factors influencing last-minute cancellations included the impact of the coronavirus disease 2019 pandemic, average wind speed, average rainfall, preanesthetic assessments, and patient age. The extreme gradient boosting algorithm outperformed other models in predicting cancellations, boasting an area under the curve value of 0.923 and an accuracy of 0.841. This algorithm yielded superior sensitivity (0.840), precision (0.837), and F1 score (0.838) relative to the other models. These insights underscore the potential of machine learning, informed by EMRs and meteorological data, in forecasting last-minute surgical cancellations. The extreme gradient boosting algorithm holds promise for clinical deployment to curtail healthcare expenses and avert adverse patient-family experiences.
RESUMEN
Deep learning techniques have been investigated for the computer-aided diagnosis of thyroid nodules in ultrasound images. However, most existing thyroid nodule detection methods were simply based on static ultrasound images, which cannot well explore spatial and temporal information following the clinical examination process. In this paper, we propose a novel video-based semi-supervised framework for ultrasound thyroid nodule detection. Especially, considering clinical examinations that need to detect thyroid nodules at the ultrasonic probe positions, we first construct an adjacent frame guided detection backbone network by using adjacent supporting reference frames. To further reduce the labour-intensive thyroid nodule annotation in ultrasound videos, we extend the video-based detection in a semi-supervised manner by using both labeled and unlabeled videos. Based on the detection consistency in sequential neighbouring frames, a pseudo label adaptation strategy is proposed for the refinement of unpredicted frames. The proposed framework is validated on 996 transverse viewed and 1088 longitudinal viewed ultrasound videos. Experimental results demonstrated the superior performance of our proposed method in the ultrasound video-based detection of thyroid nodules.
Asunto(s)
Aprendizaje Profundo , Interpretación de Imagen Asistida por Computador , Nódulo Tiroideo , Ultrasonografía , Grabación en Video , Nódulo Tiroideo/diagnóstico por imagen , Humanos , Ultrasonografía/métodos , Interpretación de Imagen Asistida por Computador/métodos , Grabación en Video/métodos , Algoritmos , Glándula Tiroides/diagnóstico por imagenRESUMEN
The purpose of this study was to investigate the risk factors for delayed chemotherapy-induced vomiting (DCIV) in pediatric oncology patients. We collected data on pediatric patients from a tertiary care pediatric hospital in an Asian urban center. We analyzed the risk factors for DCIV in patients by univariate analysis and logistic regression. Patients were grouped according to age by the Youden index, and differences in clinical features between the high-risk and low-risk groups were calculated. In the univariate analysis, the number of chemotherapy days, pH, and blood glucose levels were significantly associated with DCIV. In the logistic regression analysis, patient age was an independent risk factor (odds ratio [OR] = 1.013, 95% confidence interval [CI] = 1.005-1.021, P = .002). Children in the high-risk group had a higher grade of vomiting (P < .05). Age is an important risk factor for DCIV in pediatric patients, with older children tending to experience more frequent and more severe vomiting.
RESUMEN
The extensive use of carbon fiber-reinforced plastics (CFRP) in aerospace, civil engineering, and other fields has resulted in a significant amount of waste, leading to serious environmental issues. Finding appropriate methods for recycling CFRP waste and effectively reusing recycled carbon fibers (rCFs) has become a challenging task. This paper presents an overview of the current status of CFRP waste and provides a systematic review and analysis of recycling technologies. In addition to discussing mechanical recycling, thermal decomposition, and chemical solvent degradation methods, the organic alkali/organic solvent method for recycling resins is also elucidated. By introducing the recycling conditions and outcomes of the organic alkali/organic solvent method, the study highlights its significance as a reference for carbon fiber recycling. Furthermore, the paper reviews the current state of rCFs utilization based on its application domains, focusing on research advancements in fiber composites and cementitious composites. Based on these findings, the paper summarizes the existing research limitations and identifies specific areas that require further attention in recycling techniques and rCFs utilization. Lastly, this review provides a prospect on the future of recycling and reusing CFRP waste.
RESUMEN
The optical characteristics of multimode luminescent materials like multimode luminescence (photoluminescence, afterglow, thermoluminescence) and a multi-excitation source (light, thermal, mechanical force) play crucial roles in optical data storage and readout, document security and anticounterfeiting. A higher level of advanced anticounterfeiting may rely on multimode anticounterfeiting materials that can realize multicolor luminescence. Here, a highly integrated multimode and multicolor Y7O6F9:Er3+,Eu3+ material is developed through multiplexing of dual lanthanides in fluorine oxide particles. In photoluminescence and photoluminescence/up-conversion luminescence modes, the material Y7O6F9:Er3+,Eu3+ has the characteristic of excitation wavelength and power dependence. In the photoluminescence mode, under excitation at 254 nm and 365 nm, Y7O6F9:Er3+ and Y7O6F9:Eu3+ showed bright red and green emissions, respectively. In the photoluminescence/up-conversion mode, under the increased excitation power from 0.2 to 2.0 W cm-2, the color of luminescence emission can be finely tuned from red to orange, yellow and green. Taking this unique excitation wavelength-power-dependent luminescence property into account, a multilevel anticounterfeiting device with the Lily pattern was designed. The device readily integrates the advantages of the excitation wavelength-dependent photoluminescence emissions and excitation power-dependent photoluminescence emissions in one overall device. These findings offer unique insight for designing highly integrated multimode, multicolor luminescence materials and advanced anticounterfeiting technology toward a wide variety of applications, particularly multilevel anticounterfeiting devices.
RESUMEN
The development of portable and cost-effective sensing system for Hg2+ quantitation is highly demanded for environmental monitoring. Herein, an on-site, rapid and portable smartphone readout device based Hg2+ sensing system integrating nitrogen-doped carbon quantum dots (NCDs) modified paper strip was proposed, and the physicochemical properties of NCDs were characterized by high resolution TEM, FTIR, UV-vis absorption spectrum and fluorescence spectral analysis. The modified paper strip was prepared via "ink-jet" printing technology and exhibits sensitive fluorescence response to Hg2+ with fluorescence color of bright blue (at the excitation/emission wavelength of 365/440 nm). This portable smartphone-based sensing platform is highly selective and sensitive to Hg2+ with the limit of detection (LOD) of 10.6 nM and the concentration range of 0-130 nM. In addition, the recoveries of tap water and local lake water were in the range of 89.4% to 109%. The cost-effective sensing system based on smartphone shows a great potential for trace amounts of Hg2+ monitoring in environmental water samples.
RESUMEN
Pain management is a crucial concern in medicine, particularly in the case of children who may struggle to effectively communicate their pain. Despite the longstanding reliance on various assessment scales by medical professionals, these tools have shown limitations and subjectivity. In this paper, we present a pain assessment scheme based on skin potential signals, aiming to convert subjective pain into objective indicators for pain identification using machine learning methods. We have designed and implemented a portable non-invasive measurement device to measure skin potential signals and conducted experiments involving 623 subjects. From the experimental data, we selected 358 valid records, which were then divided into 218 silent samples and 262 pain samples. A total of 38 features were extracted from each sample, with seven features displaying superior performance in pain identification. Employing three classification algorithms, we found that the random forest algorithm achieved the highest accuracy, reaching 70.63%. While this identification rate shows promise for clinical applications, it is important to note that our results differ from state-of-the-art research, which achieved a recognition rate of 81.5%. This discrepancy arises from the fact that our pain stimuli were induced by clinical operations, making it challenging to precisely control the stimulus intensity when compared to electrical or thermal stimuli. Despite this limitation, our pain assessment scheme demonstrates significant potential in providing objective pain identification in clinical settings. Further research and refinement of the proposed approach may lead to even more accurate and reliable pain management techniques in the future.
Asunto(s)
Dolor , Piel , Humanos , Niño , Dolor/diagnóstico , Algoritmos , Aprendizaje Automático , Bosques AleatoriosRESUMEN
In recent years, a series of persistent luminescence materials excitable by blue light have been developed and widely used in many fields such as optical information storage, AC-LEDs, anti-counterfeiting and bio-imaging. However, it is still a long-standing challenge to develop a superior red-emitting persistent phosphor that can be efficiently excited by blue light. In this work, a novel blue-light excited red-emitting persistent phosphor CaCd2Ga2Ge3O12:Pr3+ was successfully synthesized by using a solid-state method, showing excellent luminescence properties. Moreover, the phase purity, crystal structure, photoluminescence spectra, afterglow emission spectra, and three-dimensional thermoluminescence spectrum were successfully investigated. Under 294 nm excitation, photoluminescence spectra show a single orange emission and a series of peaks centered at 492, 537, 568, 614 and 664 nm, which correspond to the 3P0 â 3H4, 3P0 â 3H5, 3P2 â 3H6, 1D2 â 3H4, and 3P0 â 3F2 transitions of Pr3+, respectively. Interestingly, after blue light excitation, the afterglow luminescence exhibits red long emission, which is attributed to the 1D2 â 3H4 transition of Pr3+. Through thermoluminescence spectra and three-dimensional thermoluminescence spectra, we analyze the reasons for the different colors of photoluminescence and afterglow luminescence. The results imply that there are two types of traps, and the depth of shallow traps and deep traps is calculated to be 0.684 and 0.776 eV, respectively. It is worth noting that the photoluminescence is attributed to the 4f2 â 4f5d and f â f transitions of Pr3+, and the afterglow luminescence is ascribed to a tunneling-related process and the transition of electrons from the valence band to the conduction band. The obtained red-emitting persistent phosphors provide a promising pathway toward AC-LEDs, multi-cycle bio-imaging and other fields.
RESUMEN
INTRODUCTION: Pulmonary rehabilitation (PR) is a comprehensive nursing intervention for lung function improvement in patients with respiratory diseases. This systematic review focused on further exploration of the unclear impacts of PR on childhood asthma. EVIDENCE ACQUISITION: Web of Science, Cochrane Library, Embase, PubMed, and other databases were searched until May 2021. Randomized controlled trials (RCTs) comparing the effects of PR (including exercise training and education) and routine care on childhood asthma were included. Study selection, data extraction, and bias risk assessment were performed independently by two investigators. EVIDENCE SYNTHESIS: Fourteen RCTs involving 1401 patients were included. Relative to the control group, the total scores of the asthma quality of life questionnaire were evidently improved in the experimental group, including motor domain scores (MD=0.88, 95% CI: 0.67-1.09), symptom domain scores (MD=1.23, 95% CI: 0.61-1.85), and affective domain scores (MD=1.38, 95% CI: 0.63-2.14). Besides, 6-min walk distance (MD=2.01, 95% CI: 0.86-3.15) and asthma control test (MD=0.31, 95% CI: 0.02-0.60) were prominently improved. However, the maximum oxygen uptake (MD=0.81, 95% CI: -0.2 to 1.82) was not markedly improved. The forced expiratory volume in 1 s (MD=0.42, 95% CI: -0.29 to 1.13), and forced vital capacity (MD=0.07, 95% CI: -0.14 to 0.28) were not remarkably improved. There was an evident improvement in the peak expiratory flow (MD=1.22, 95% CI: 0.15-2.30). CONCLUSIONS: PR improves some lung functions, exercise tolerance and the quality of life of children with asthma, and it might work as a supplementary therapy for the treatment of childhood asthma. Moreover, more RCTs of high quality and in large sample size are needed for further confirmation.
Asunto(s)
Asma , Niño , Humanos , Asma/terapia , Calidad de Vida , Ejercicio Físico , Caminata , Volumen Espiratorio ForzadoRESUMEN
OBJECTIVE: The objective of this study was to appraise the interrelation between overweight/obesity and the safety and efficacy of COVID-19 vaccination by synthesizing the currently available evidence. METHODS: A systematic review of published studies on the safety and efficacy of the COVID-19 vaccine in people who were overweight or obese was conducted. Databases including Embase, Medline Epub (Ovid), PsychInfo (Ovid), Web of Science, PubMed, CINAHL, and Google Scholar were searched to identify relevant studies. The databases of the Centers for Disease Control (CDC) and World Health Organization (WHO) were also searched for relevant unpublished and gray literature. RESULTS: Fifteen studies were included in the review. All the included studies used observational study designs; there were ten cohort studies and five cross-sectional studies. The sample size of these studies ranged from 21 to 9,171,524. Thirteen studies reported using BNT162b2 (Pfizer-BioNTech, USA), four reported using ChAdOx-nCov19 (AstraZeneca, U.K), two were reported using CoronaVac (Sinovac, China), and two were reported using mRNA1273 (Moderna, USA). The efficacy and safety of COVID-19 vaccines have been extensively studied in individuals with overweight/obesity. Most studies have shown that the humoral response decreases with increasing BMI. The available evidence does not conclusively indicate that these vaccines are generally safe in this population. CONCLUSION: While the efficacy of the COVID-19 vaccine may be less than ideal in people who are overweight or obese, it does not mean that obese people should not be vaccinated, as the vaccine can still provide some protection. There is a lack of evidence for conclusions to be drawn about the safety of the vaccine in the population. This study calls on health professionals, policymakers, caregivers, and all other stakeholders to focus on monitoring the possible adverse effects of injections in overweight/obese people.
RESUMEN
Recently, tremendous human-designed and automatically searched neural networks have been applied to image denoising. However, previous works intend to handle all noisy images in a pre-defined static network architecture, which inevitably leads to high computational complexity for good denoising quality. Here, we present a dynamic slimmable denoising network (DDS-Net), a general method to achieve good denoising quality with less computational complexity, via dynamically adjusting the channel configurations of networks at test time with respect to different noisy images. Our DDS-Net is empowered with the ability of dynamic inference by a dynamic gate, which can predictively adjust the channel configuration of networks with negligible extra computation cost. To ensure the performance of each candidate sub-network and the fairness of the dynamic gate, we propose a three-stage optimization scheme. In the first stage, we train a weight-shared slimmable super network. In the second stage, we evaluate the trained slimmable super network in an iterative way and progressively tailor the channel numbers of each layer with minimal denoising quality drop. By a single pass, we can obtain several sub-networks with good performance under different channel configurations. In the last stage, we identify easy and hard samples in an online way and train a dynamic gate to predictively select the corresponding sub-network with respect to different noisy images. Extensive experiments demonstrate our DDS-Net consistently outperforms the state-of-the-art individually trained static denoising networks.
RESUMEN
Recognition and quantitative analytics of histopathological cells are the golden standard for diagnosing multiple cancers. Despite recent advances in deep learning techniques that have been widely investigated for the automated segmentation of various types of histopathological cells, the heavy dependency on specific histopathological image types with sufficient supervised annotations, as well as the limited access to clinical data in hospitals, still pose significant challenges in the application of computer-aided diagnosis in pathology. In this paper, we focus on the model generalization of cell segmentation towards cross-tissue histopathological images. Remarkably, a novel target-specific finetuning-based self-supervised domain adaptation framework is proposed to transfer the cell segmentation model to unlabeled target datasets, without access to source datasets and annotations. When performed on the target unlabeled histopathological image set, the proposed method only needs to tune very few parameters of the pre-trained model in a self-supervised manner. Considering the morphological properties of pathological cells, we introduce two constraint terms at both local and global levels into this framework to access more reliable predictions. The proposed cross-domain framework is validated on three different types of histopathological tissues, showing promising performance in self-supervised cell segmentation. Additionally, the whole framework can be further applied to clinical tools in pathology without accessing the original training image data. The code and dataset are released at: https://github.com/NeuronXJTU/SFDA-CellSeg.
Asunto(s)
Diagnóstico por Computador , Procesamiento de Imagen Asistido por Computador , Aprendizaje Automático SupervisadoRESUMEN
BACKGROUND: Exploring the experience and understanding of death in children with terminal cancer is important to provide them with appropriate care. However, most studies have focused on the perspectives of parents and healthcare professionals, and few have focused on the end-of-life experiences of children. AIM: To advance the understanding of end-of-life experiences and perceptions of death in children with cancer. DESIGN: Interpretative qualitative study using semi-structured interviews. Data were analyzed using reflexive thematic analysis. SETTING/PARTICIPANTS: The study was conducted at the department of oncological surgery, Children's Hospital, Zhejiang University School of Medicine. Ten children aged 8-17 with terminal cancer were included in the study. RESULTS: Four major themes (and eight sub-themes) were identified from the findings: (1) helplessness in the face of death (loneliness, loss of control); (2) desire to connect with the world they left (reluctantly to be forgotten, sense of self-worth); (3) perceptions and attitudes toward death (separating from loved ones, embracing death); (4) expectations of future life (promoting comfort, fulfilling wishes). CONCLUSIONS: Children with terminal cancer have a strong sense of loneliness and a desire to connect with the world they have left behind. Differences in children's perceptions and attitudes about death suggest that healthcare professionals should focus on their experiences and needs and provide personalized palliative care services to children and their families to improve their quality of life.
Asunto(s)
Neoplasias , Cuidado Terminal , Humanos , Niño , Calidad de Vida , Padres , Investigación Cualitativa , Muerte , Cuidados PaliativosRESUMEN
Hypochlorous acid (HClO) plays an essential role in biological systems. The characteristics of potent oxidization and short lifetime make it challenging to detect specifically from other reactive oxygen species (ROS) at cellular levels. Therefore, its detection and imaging with high selectivity and sensitivity are of great significance. Herein a turn-on HClO fluorescent probe (named RNB-OCl) with boronate ester as the recognition site was designed and synthesized. The RNB-OCl displayed good selective and ultrasensitive to HClO with a low detection limit of 1.36 nM by the intramolecular charge transfer (ICT)-fluorescence resonance energy transfer (FRET) dual mechanism in reducing the fluorescence background and improving the sensitivity. In addition, the role of the ICT-FRET was further demonstrated by time-dependent density functional theory (TD-DFT) calculations. Furthermore, the probe RNB-OCl was successfully employed for imaging HClO in living cells.
Asunto(s)
Transferencia Resonante de Energía de Fluorescencia , Ácido Hipocloroso , Animales , Ratones , Colorantes Fluorescentes , Imagen Óptica/métodos , Células RAW 264.7RESUMEN
BACKGROUND: Effective family management can improve the ability of children's families to cope with the disease. However, the family management styles of children with leukemia are unknown. OBJECTIVE: This study used a family management style framework and structural equation model to determine the direct and indirect predictors of family management of children with leukemia. METHODS: A total of 496 parents of children with leukemia completed the Family Management Measure, the Feetham Family Functioning Survey, the Social Support Rating Scale, and a demographic family information form. We used structural equation modeling to explore the direct and indirect effects of the study variables on family management styles. RESULTS: Family functioning, family income, parental education level, and geographic location affect family management directly, among which family functioning has the most significant coefficient in the model (0.53). Siblings and social support can affect family management through the mediating effect of family functioning. Employment change can directly affect family management (-0.27) and indirectly affect family functioning (-0.08). CONCLUSIONS: Our results support the family management style framework. Individual and family factors such as family income, parental education level, geographic location, employment change, siblings, and social support contribute to the ease or difficulty of family management. In addition, family functioning can also mediate between contextual influences and family management styles. IMPLICATIONS FOR PRACTICE: Family functioning plays a vital role in relationships between contextual influences and family management. We can construct family management interventions according to family management influencing factors.
Asunto(s)
Pueblos del Este de Asia , Leucemia , Niño , Humanos , Leucemia/terapia , Padres , Hermanos , Apoyo SocialRESUMEN
Generally, pulmonary artery sling operation involves the pulmonary artery transplantation to be cut off. Nursing care is focused on the postoperative pulmonary vascular anastomosis, respiratory tract, and blood pressure after surgery. We report the case of an infant who underwent pulmonary artery tracheal transposition combined with Slide keratoplasty, where the pulmonary artery transplantation was not cut off. We highlight that postoperative pulmonary artery blood flow to the unobstructed airway and airway reconstruction surgery should be focused on to help children recover and ensure successful surgery. This case study reports the postoperative nursing experience of one patient with pulmonary artery sling undergoing pulmonary tracheal transposition combined with Slide arthroplasty. This demonstrates that throughout the postoperative care, airway management should be focused on to maintain circulation stability in the early postoperative period, and corresponding measures such as posture management, atomization inhalation, and improved chest physical therapy should be applied according to the special surgical method of the case in order to reduce airway complications and to improve the surgical success rate of children with pulmonary artery sling undergoing pulmonary tracheal transposition combined with Slide arthroplasty. In similar cases, after pulmonary tracheal transposition and Slide angioplasty, the doctors and nurses should pay attention to early circulation stability and focus on airway management through careful treatment and nursing, so as to promote the child's recovery.
Asunto(s)
Cardiopatías Congénitas , Estenosis Traqueal , Lactante , Niño , Humanos , Arteria Pulmonar/cirugía , Resultado del Tratamiento , Cuidados Posoperatorios , Estudios Retrospectivos , Cardiopatías Congénitas/cirugíaRESUMEN
OBJECTIVE: To describe the temporal trend of the number of new congenital heart disease (CHD) cases among newborns in Jinhua from 2019 to 2020 and explored an appropriate model to fit and forecast the tendency of CHD. METHODS: Data on CHD from 2019 to 2020 was collected from a health information system. We counted the number of newborns with CHD weekly and separately used the additive Holt-Winters ES method and ARIMA model to fit and predict the number of CHD for newborns in Jinhua. By comparing the mean square error, rooted mean square error and mean absolute percentage error of each approach, we evaluated the effects of different approaches for predicting the number of CHD in newborns. RESULTS: A total of 1135 newborns, including 601 baby girls and 534 baby boys, were admitted for CHD from HIS in Jinhua during the 2-year study period. The prevalence of CHD among newborns in Jinhua in 2019 was 0.96%. Atrial septal defect was diagnosed the most frequently among all newborns with CHD. The number of CHD cases among newborns remained stable in 2019 and 2020. There were fewer cases in spring and summer, while cases peaked in November and December. The ARIMA(2,1,1) model relatively offered advantages over the additive Holt-winters ES method in predicting the number of newborns with CHD, while the accuracy of ARIMA(2,1,1) was not very ideal. CONCLUSIONS: The diagnosis of CHD is related to many risk factors, therefore, when using temporal models to fit and predict the data, we must consider such factors' influence and try to incorporate them into the models.
Asunto(s)
Cardiopatías Congénitas , Proyectos de Investigación , Predicción , Cardiopatías Congénitas/diagnóstico , Cardiopatías Congénitas/epidemiología , Humanos , Recién Nacido , Modelos Estadísticos , Factores de Riesgo , Estaciones del AñoRESUMEN
OBJECTIVE: To understand the risk factors associated with adverse events during exchange transfusion (ET) in severe neonatal hyperbilirubinemia. STUDY DESIGN: We conducted a retrospective study of infants with hyperbilirubinemia who underwent ET within 30 days of birth from 2015 to 2020 in a children's hospital. Both traditional statistical analysis and state-of-the-art explainable artificial intelligence (XAI) were used to identify the risk factors. RESULTS: A total of 188 ET cases were included; 7 major adverse events, including hyperglycemia (86.2%), top-up transfusion after ET (50.5%), hypocalcemia (42.6%), hyponatremia (42.6%), thrombocytopenia (38.3%), metabolic acidosis (25.5%), and hypokalemia (25.5%), and their risk factors were identified. Some novel and interesting findings were identified by XAI. CONCLUSIONS: XAI not only achieved better performance in predicting adverse events during ET but also helped clinicians to more deeply understand nonlinear relationships and generate actionable knowledge for practice.