ABSTRACT
Background- This cross sectional study was conducted to correlate the various smile characteristics such as- smile line, smile arcs, lip curvature, gingival exposure and maxillary central incisor exposure with gender in north Indian population. The study was conducted in department of Oral Pathology and Microbiology department at Shree BankeyMaterials and methods- Bihari Dental College & Research Centre, Uttar Pradesh, India. A total of 100 research participants were selected according to inclusion and exclusion criteria out of which 56 were males and 44 were females. To obtain the smile photographs, constant camera settings was used. The collected data were analyzed using Image-J software for sexual dimorphism and forensic identi?cation. Result-Results of this study showed, out of 100 individuals 73 individuals have average smile line followed by low smile line & high smile line. Out of 100 individuals, 67 individuals have have parallel smile arcs followed by ?ate smile arcs & reverse smile arcs. Out of 100 individuals 54 individuals have straight upper lip curvature followed by upward & downward upper lip curvature. Out of 100 individuals 48 have inter proximal gingival exposure followed by no gingival exposure & gingival exposure. Out of 100 individuals 57 individuals have 75% central incisor exposure. Out of 100 individuals 41 individuals showed square shaped incisors followed by rectangular shaped, round shape & triangular shape. Each individual's facial and dental features are unique, making them valuable in forensic dentistry for Conclusion- identi?cation. These distinct traits also play a crucial role in enhancing personalized treatments in aesthetic dentistry.
ABSTRACT
Background: Pre-eclampsia is a multi-system, pregnancy specific disorder that is characterized by the development of hypertension and proteinuria after 20 weeks. Pre-eclampsia is the majority of referrals to tertiary care centre. It is one of the major causes of maternal and perinatal morbidity and mortality.Methods: A retrospective analytical study done over a period of six months from January 1st 2023 to June 30th 2023. Pregnant women admitted with PE with severe features to Cheluvamba hospital, MMCRI, Mysore during the study were considered and analysed using the proforma. Data was entered into Microsoft excel data sheet and was analyzed. Categorical data was represented in the form of Frequencies and proportionsResults: Incidence of PE with severe features in our hospital was 3.4%. Majority (69%) were between 23-27 years of age and 52.7% were primigravida. Maternal complications were noted in 37.5% attributed to renal dysfunction, postpartum haemorrhage, DIC, placental abruption, HELLP, pulmonary oedema and postpartum eclampsia.Conclusions: Maternal and perinatal complications are more in patients with severe pre-eclampsia. The incidence of severe pre-eclampsia can be reduced by early referral, better antenatal care, early recognition and treatment of pre-eclampsia
ABSTRACT
A high level of disorder in many viral proteins is a direct consequence of their small genomes, which makes interaction with multiple binding partners a necessity for infection and pathogenicity. A segment of the flaviviral capsid protein (C), also known as the molecular recognition feature (MoRF), undergoes a disorder-to-order transition upon binding to several protein partners. To understand their role in pathogenesis, MoRFs were identified and their occurrence across different flaviviral capsids were studied. Despite lack of sequence similarities, docking studies of Cs with the host proteins indicate conserved interactions involving MoRFs across members of phylogenetic subclades. Additionally, it was observed from the protein–protein networks that some MoRFs preferentially bind proteins that are involved in specialized functions such as ribosome biogenesis. The findings point to the importance of MoRFs in the flaviviral life cycle, with important consequences for disease progression and suppression of the host immune system. Potentially, they might have impacted the way flaviviruses evolved to infect varied hosts using multiple vectors.
ABSTRACT
Objective To analyze the echinococcosis surveillance results in Bayingolin Mongol Autonomous Prefecture, Xinjiang Uygur Autonomous Region from 2017 to 2022, so as to provide insights into formulation of echinococcosis control measures in the prefecture. Methods Villagers were randomly sampled using a multistage sampling method from class I and II echinococcosis endemic counties in Bayingolin Mongolian Autonomous Prefecture from 2017 to 2022 for detection of human echinococcosis, while all patients undergoing ultrasound examinations in medical institutions in class III endemic counties received active echinococcosis screening. In addition, livestock in centralized slaughterhouses or slaughtering sites were screened for echinococcosis using the palpation and necropsy method, and fresh domestic dog feces samples were collected from randomly selected dog owners in each administrative village for detection of Echinococcus copro-antigen in domestic dogs. The trends in detection of human and livestock echinococcosis, detection of newly diagnosed human echinococcosis cases and detection of Echinococcus coproantigen in domestic dogs were analyzed in Bayingolin Mongol Autonomous Prefecture from 2017 to 2022. Results The mean detection rate of human echinococcosis was 0.13% (540/407 803) in Bayingolin Mongol Autonomous Prefecture from 2017 to 2022, which appeared a tendency towards a decline over years (χ2trend = 1 217.21, P < 0.001), and the highest detection of newly diagnosed echinococcosis cases was seen in Hejing County (0.28%, 191/67 865). The detection of livestock echinococcosis appeared a tendency towards a decline over years from 2017 to 2022 (χ2trend = 147.02, P < 0.001), with the highest detection rate seen in Hejing County (3.44%, 86/2 500), and the detection of Echinococcus copro-antigen in domestic dogs appeared a tendency towards a decline over years from 2017 to 2022 (χ2trend = 302.46, P < 0.001), with the highest detection rate in Qiemo County (2.74%, 118/4 313). Conclusions The detection of human and livestock echinococcosis and dog feces antigens Echinococcus copro-antigen in domestic dogs all appeared a tendency towards a decline in Bayingolin Mongol Autonomous Prefecture, Xinjiang Uygur Autonomous Region from 2017 to 2022; however, there is still a high echinococcosis transmission risk in local areas. Sustainable integrated echinococcosis control is required in Bayingolin Mongol Autonomous Prefecture.
ABSTRACT
The treatment of glioblastoma, the most prevalent malignant tumor in the central nervous system, poses considerable challenges. Glioblastoma multiforme, classified as a grade Ⅳ highly malignant brain glioma by the World Health Organization, is typically managed through a combination of surgery, postoperative chemotherapy, and radiotherapy. The treatment of glioblastoma is complicated by its infiltrative nature, genetic heterogeneity, and presence of the blood-brain barrier. Almost all cases of glioblastoma experience recurrence despite aggressive therapy, exploring the development of updated molecular treatment strategies that can improve overall efficacy. A crucial aspect in modern neurosurgery is the precise delineation of brain regions in terms of their anatomy and function. It serves as the fundamental basis for investigating variations in the distribution of brain gliomas. Hence, this review will elucidate the origin of glioblastomas and analyze the potential factors contributing to the spatially specific distribution of gliomas on the basis of a theoretical framework of brain connectomics research. Molecular characteristics, information pathways, tumor microenvironment landscape, and immunology will inform the analysis. We aim to identify novel biomolecular targets and therapeutic pathways to gain scientific insights for effective glioblastoma treatment.
ABSTRACT
ObjectiveTo explore the clinical features and causative genes of short stature children with unknown etiology， providing evidence for precise clinical diagnosis and treatment. MethodsThe study recruited children with suspected but undiagnosed short stature from the pediatric endocrinology department in our hospital between January 2018 and August 2022. A retrospective analysis was performed on the clinical manifestations， laboratory test and whole exome sequencing （WES） results. Causative genes were classified and analyzed according to different pathogenic mechanisms. ResultsA total of 48 children （30 boys and 18 girls） were enrolled， aged 7.73 ± 3.97 years， with a height standard deviation score （ HtSDS） of -3.63 ± 1.67. Of the patients， 33 （68.8%） suffered from facial anomalies， 31 （64.6%） from skeletal abnormalities， 26 ［54.2%， 61.5% of whom born small for gestational age （SGA）］ from perinatal abnormalities， 24 ［50.0%， 87.5% of whom with growth hormone （GH） peak concentration below normal］ from endocrine disorders and 21（43.8%） had a family history of short stature. Laboratory tests showed that GH peak concentration following stimulation test was （9.72 ± 7.25） ng/mL， IGF-1 standard deviation score was -0.82 ± 1.42， the difference between bone age and chronological age was -0.93 ± 1.39 years. Of the 25 cases with mutant genes found by WES， 14 （56.0%） had pathogenic mutation， 6 （24.0%） likely pathogenic mutation， and 5 （20.0%） mutation of uncertain significance. Pathogenic and likely pathogenic variants were identified in 14 genes， including 10 affecting intracellular signaling pathways （PTPN11， RAF1， RIT1， ARID1B， ANKRD11， CSNK2A1， SRCAP， CUL7， SMAD4 and FAM111A） and 4 affecting extracellular matrix （ECM） components or functions （ACAN， FBN1， COL10A1 and COMP）. ConclusionsA rare monogenic disease should be considered as the possible etiology for children with severe short stature accompanied by facial anomalies， disproportionate body types， skeletal abnormalities， SGA， GH peak concentration below normal and a family history of short stature. WES played an important role in identifying the monogenic causes of short stature. This study indicated that affecting growth plate cartilage formation through intracellular signaling pathways and ECM components or functions was the main mechanism of causative genes leading to severe short stature in children. Further research may help discover and study new pathogenic variants and gene functions.
ABSTRACT
Video-based intelligent action recognition remains challenging in the field of computer vision.The review analyzes the state-of-the-art methods of video-based intelligent action recognition,including machine learning methods with handcrafted features,deep learning methods with automatically extracted features,and multi-information fusion methods.In addition,the important medical applications and limitations of these technologies in the past decade are introduced,and the interdisciplinary views on the future application to improve human health are also shared.
ABSTRACT
Objective To explore the predictive value of CT radiomics and morphological features for the prognosis and survival in non-small cell lung cancer(NSCLC)patients.Methods The clinic data of 300 NSCLC patients(300 lesions)were downloaded from the Cancer Imaging Archive,with 210 randomly selected as the training set and 90 as the test set.According to the prognosis and survival,the patients were divided into two groups with survival period≤3 and＞3 years.3D Slicer software was used to delineate the regions of interest layer by layer in CT images,and the radiomics features were extracted from each region of interest.Both t-test and least absolute shrinkage and selection operator were utilized for radiomics feature screening.Three types of prediction models,namely radiomics model,morphological model and combined model,were constructed with Logistic regression,whose performances were evaluated using the receiver operating characteristic(ROC)curve.Results The differences in radiomics labels and mediastinal lymph node metastasis between the training set and the test set were statistically significant.For radiomics model,morphological model and combined model,the area under the ROC curve was 0.784(95%CI:0.722-0.847),0.734(95%CI:0.664-0.804)and 0.748(95%CI:0.680-0.815)in the training set,and 0.737(95%CI:0.630-0.844),0.665(95%CI:0.554-0.777)and 0.687(95%CI:0.578-0.797)in the test set,which demonstrated that radiomics model had the best diagnostic performance.Conclusion The CT radiomics model can effectively predict the prognosis and survival in NSCLC patients.
ABSTRACT
A U-Net incorporating improved Transformer and convolutional channel attention module is designed for biventricular segmentation in MRI image.By replacing the high-level convolution of U-Net with the improved Transformer,the global feature information can be effectively extracted to cope with the challenge of poor segmentation performance due to the complex morphological variation of the right ventricle.The improved Transformer incorporates a fixed window attention for position localization in the self-attention module,and aggregates the output feature map for reducing the feature map size;and the network learning capability is improved by increasing network depth through the adjustment of multilayer perceptron.To solve the problem of unsatisfactory segmentation performance caused by blurred tissue edges,a feature aggregation module is used for the fusion of multi-level underlying features,and a convolutional channel attention module is adopted to rescale the underlying features to achieve adaptive learning of feature weights.In addition,a plug-and-play feature enhancement module is integrated to improve the segmentation performance which is affected by feature loss due to channel decay in the codec structure,which guarantees the spatial information while increasing the proportion of useful channel information.The test on the ACDC dataset shows that the proposed method has higher biventricular segmentation accuracy,especially for the right ventricle segmentation.Compared with other methods,the proposed method improves the DSC coefficient by at least 2.83%,proving its effectiveness in biventricular segmentation.
ABSTRACT
In response to the challenges of varied sizes and diverse shapes of colorectal polyps,especially with blurred boundaries that often complicates localization and smaller polyps being particularly prone to oversight,a colorectal polyp segmentation algorithm integrating Transformer and convolution is proposed.Transformer is employed to extract global features from images for ensuring the network's capability for global modeling and improving the localization capability for both main polyp regions and vague boundaries.Subsequently,convolution is introduced to augment the network's ability to process polyp details,refining boundary segmentation and enhancing the capture capability for small-sized polyps.Finally,a deep fusion of the features extracted by Transformer and convolution is carried out to realize feature complementarity.The experimental evaluation using CVC-ClinicDB and Kvasir-SEG datasets show that the algorithm has similarity coefficients of 95.4%and 93.2%,and mean intersection over union of 91.3%and 88.6%,respectively.Further tests on the generalization capability of the algorithm are conducted on CVC-ColonDB,CVC-T,and ETIS datasets,in which similarity coefficients of 81.3%,90.9%and 80.1%are obtained.The results indicate a notable improvement in the accuracy of polyp segmentation achieved by the proposed algorithm.
ABSTRACT
To address the problem of low accuracy in multi-classification recognition of motor imagery electroencephalogram(EEG)signals,a recognition method is proposed based on differential entropy and convolutional neural network for 4-class classification of motor imagery.EEG signals are extracted into 4 frequency bands(Alpha,Beta,Theta,and Gamma)through the filter,followed by the computation of differential entropy for each frequency band.According to the spatial characteristics of brain electrodes,the data structure is reconstructed into three-dimensional EEG signal feature cube which is input into convolutional neural network for 4-class classification.The method achieves an accuracy of 95.88%on the BCI Competition IV-2a public dataset.Additionally,a 4-class classification motor imagery dataset is established in the laboratory for the same processing,and an accuracy of 94.50%is obtained.The test results demonstrate that the proposed method exhibits superior recognition performance.
ABSTRACT
Tumors are serious diseases threatening human health,and the early diagnosis is essential to improve treatment success and patient survival.The study of tumor gene expression data has become a major tool for revealing tumor disease mechanisms,in which artificial intelligence plays an important role.The potential advantages of supervised learning,unsupervised learning and deep learning in tumor prediction and classification are explored from the perspective of machine learning methods.Special attention is paid to the impact of feature selection algorithms on gene screening and their importance in high-dimensional gene expression data.By providing a comprehensive overview of the application and development of artificial intelligence in the analysis of tumor gene expression data,the study aims to provide an outlook for future research directions and promote further development.
ABSTRACT
Objective To explore the method of objective identification of color information in sublingual veins diagnosis of TCM.Methods Combined with computer vision,compact fully convolution networks(CFCNs)and 19 deep learning classification models were used for study,and a double pulse rectangle algorithm was designed as a means of segmentation and recognition of sublingual veins and color information extraction.Results The accuracy of segmentation of tongue bottom obtained by the method of removing reflection + data expanding + data post-processing was 0.955 9,F1 value was 0.947 3,and mIoU value was 0.900 0.The accuracy of segmentation of sublingual veins obtained by the method of removing reflection + tongue input + data expanding + corrosion expansion was 0.778 4,F1 value was 0.738 3 and mIoU value was 0.585 1,which were obviously superior to the current classic or improved U-net model.On the color classification of sublingual veins,the best classification model was DenseNet161-bc-early_stopping with an accuracy rate of 0.803 7.Conclusion The deep learning method has a certain effect on identifying the color information of sublingual veins in TCM,which provides a new method for the research of quantitative color detection technology of sublingual veins diagnosis in TCM.
ABSTRACT
@#Objective To explore the correlation between the quantitative and qualitative features of CT images and the invasiveness of pulmonary ground-glass nodules, providing reference value for preoperative planning of patients with ground-glass nodules. Methods The patients with ground-glass nodules who underwent surgical treatment and were diagnosed with pulmonary adenocarcinoma from September 2020 to July 2022 at the Third Affiliated Hospital of Kunming Medical University were collected. Based on the pathological diagnosis results, they were divided into two groups: a non-invasive adenocarcinoma group with in situ and minimally invasive adenocarcinoma, and an invasive adenocarcinoma group. Imaging features were collected, and a univariate logistic regression analysis was conducted on the clinical and imaging data of the patients. Variables with statistical difference were selected for multivariate logistic regression analysis to establish a predictive model of invasive adenocarcinoma based on independent risk factors. Finally, the sensitivity and specificity were calculated based on the Youden index. Results A total of 555 patients were collected. The were 310 patients in the non-invasive adenocarcinoma group, including 235 females and 75 males, with a meadian age of 49 (43, 58) years, and 245 patients in the invasive adenocarcinoma group, including 163 females and 82 males, with a meadian age of 53 (46, 61) years. The binary logistic regression analysis showed that the maximum diameter (OR=4.707, 95%CI 2.060 to 10.758), consolidation/tumor ratio (CTR, OR=1.027, 95%CI 1.011 to 1.043), maximum CT value (OR=1.025, 95%CI 1.004 to 1.047), mean CT value (OR=1.035, 95%CI 1.008 to 1.063), spiculation sign (OR=2.055, 95%CI 1.148 to 3.679), and vascular convergence sign (OR=2.508, 95%CI 1.345 to 4.676) were independent risk factors for the occurrence of invasive adenocarcinoma (P<0.05). Based on the independent predictive factors, a predictive model of invasive adenocarcinoma was constructed. The formula for the model prediction was: Logit(P)=–1.293+1.549×maximum diameter of lesion+0.026×CTR+0.025×maximum CT value+0.034×mean CT value+0.72×spiculation sign+0.919×vascular convergence sign. The area under the receiver operating characteristic curve of the model was 0.910 (95%CI 0.885 to 0.934), indicating that the model had good discrimination ability. The calibration curve showed that the predictive model had good calibration, and the decision analysis curve showed that the model had good clinical utility. Conclusion The predictive model combining quantitative and qualitative features of CT has a good predictive ability for the invasiveness of ground-glass nodules. Its predictive performance is higher than any single indicator.
ABSTRACT
Objective The objective of this study is to improve the accuracy of automatic identification in complex background herbal slice images.The goal is to achieve accurate recognition of herbal slice images in the presence of complex backgrounds.Methods The experiment was conducted on a collected and organized dataset of Tibetan herbal slice images.The RGB,HOG,and LBP features of the slices were analyzed.An improved HOG algorithm was used to fuse multiple features,and a deep learning network was utilized for image recognition.Results The proposed method of multi-feature fusion combined with deep learning achieved an identification accuracy of 91.68%on 3610 Tibetan herbal slice images with complex backgrounds.Furthermore,the average identification accuracy for 20 common traditional Chinese medicine slices,such as Chuan Beimu,Hawthorn,and Pinellia,reached 98.00%.This method outperformed existing methods for identifying herbal slices in complex backgrounds,indicating its feasibility and wide applicability for the identification of other traditional Chinese herbal medicines.Conclusion The fusion of multiple features effectively captures distinguishing characteristics of herbal slices in complex backgrounds.It exhibits high recognition rates for Tibetan herbal slices with complex and heavily occluded backgrounds,and can be successfully applied to the recognition of natural scene-based traditional Chinese herbal medicines and herbal slices.This approach shows promising prospects for practical applications.
ABSTRACT
BACKGROUND:Previous brain studies have mostly focused on adults and fetuses,and the developmental characteristics of young children's brainstems have rarely been studied. OBJECTIVE:To observe the brainstem development characteristics of healthy young children and to explore the age-related differences and their correlation with sex. METHODS:From January 2019 to April 2022,a retrospective study of 3.0T MRI images of 174 children aged 2 to 6 years in the Affiliated Hospital of Inner Mongolia Medical University was conducted,and the median sagittal diameter,area and angle of the brainstem(including midbrain,pons and medulla oblongata)were measured. RESULTS AND CONCLUSION:There is an age-related increase in the anterior and posterior diameters of the midbrain,pons and medulla oblongata in the 2-5 years old group as well as in the longitudinal diameter and area of the midbrain,pons and medulla oblongata in the 2-6 years old group.Except for the longitudinal diameter of the medulla oblongata,all others show a positive correlation with age(r>0,P<0.05).In the 2-3 years old group and 4-5 years old group,the children are in the rapid growth and development stage,and these two age groups can be used as the key observation indicators for the development of young children.The anterior-posterior diameter,longitudinal diameter,area of the pons and total brainstem area are strongly correlated with age,which can be used as the key observation indicators for the brainstem development in young children.
ABSTRACT
BACKGROUND:Oxidative stress is closely associated with the occurrence and progression of intervertebral disc degeneration,but its underlying mechanisms and effective treatment methods remain unclear. OBJECTIVE:To identify key genes associated with intervertebral disc degeneration accompanied by oxidative stress based on bioinformatics and three machine learning algorithms,as well as to conduct an immune infiltration analysis,followed by experimental validation. METHODS:Gene expression profiles related to intervertebral disc degeneration were obtained from the GEO database and oxidative stress-related genes obtained from the GeneCards database.Differential analysis and weighted gene co-expression networks analysis were performed on the intervertebral disc degeneration dataset.The intersection of the two analyses and the intersection with the oxidative stress-related genes were taken to obtain candidate hub genes.Gene ontology and Kyoto Encyclopedia of Genes and Genomes analyses on the candidate hub genes were performed.Machine learning algorithms(LASSO regression,SVM-RFE,and random forest)were used to select the optimal feature genes and perform the receiver operator characteristic curve validation.Simultaneously,immune infiltration analysis was conducted.Nucleus pulposus samples from patients with cervical spondylosis who were treated at the Second Hospital of Shanxi Medical University from July to November 2023 were enrolled as the intervertebral disc degeneration group and nucleus pulposus samples from patients with cervical spinal cord injury as the control group.The relative expression of feature genes in the degenerated intervertebral disc was validated using qPCR method. RESULTS AND CONCLUSION:After differential gene analysis,424 differentially expressed genes were obtained.Weighted gene co-expression networks analysis yielded 5 087 genes,and 1 399 oxidative stress genes were identified,leading to the identification of 23 candidate hub genes.Gene ontology analysis revealed that these candidate hub genes are primarily involved in bacterial defense response,molecular response to bacteria,and other biological processes.In terms of cellular component,they are associated with secretion granule lumen and cytoplasmic vesicle lumen,among others.As for molecular function,they are related to endopeptidase activity and compound binding,including sulfur compounds.Kyoto Encyclopedia of Genes and Genomes analysis demonstrated that these candidate hub genes are associated with neutrophil extracellular trap formation and the renin-angiotensin system pathway,among other signaling pathways.By applying three machine learning algorithms and conducting the receiver operator characteristic curve validation,two key genes,HSPA6 and PKD1,were determined.Immune infiltration analysis revealed a strong correlation between HSPA6 and activated dendritic cells(r=0.88,P＜0.001)as well as activated CD4+ T cells(r=-0.72,P＜0.01).Similarly,PKD1 showed close associations with effector memory CD8+ T cells(r=0.55,P＜0.05)and activated dendritic cells(r=-0.56,P＜0.05).qPCR experimental results indicated that the expression level of HSPA6 was lower in the intervertebral disc degeneration group compared with the control group(P＜0.000 1),while the expression level of PKD1 was higher in the intervertebral disc degeneration group(P＜0.000 1).These findings suggest that HSPA6 and PKD1 can serve as biomarkers for intervertebral disc degeneration accompanied by oxidative stress.Interventions targeting HSPA6 and PKD1 may hold promise for improving intervertebral disc degeneration.
ABSTRACT
BACKGROUND:The assessment of asymmetric gait quality plays a pivotal role in guiding rehabilitation training;however,the link between gait quality and kinematic-kinetic gait parameters remains ambiguous. OBJECTIVE:To formulate a machine-learning model for evaluating gait quality based on gait parameters,identify factors sensitive to gait quality from asymmetric gait parameters,investigate the relationship between gait indicators and gait quality,and provide guidance for asymmetric gait training and rehabilitation. METHODS:An asymmetric gait database was established through the creation of asymmetric conditions.Kinematic and kinetic data were collected from 8 young and 8 elderly subjects(all male,right dominant population)during gait tests.Gait quality for each test data set was assessed using symmetry indices,resulting in the creation of a gait parameter-gait quality dataset.Utilizing the Random Forest algorithm,a gait quality evaluation model was developed and key quality parameter factors were identified through differential analysis.This model was iteratively refined.The model's performance was evaluated through 10-fold cross-validation,and its effectiveness was verified using the cross-validation dataset. RESULTS AND CONCLUSION:(1)A gradient test was designed to categorize gait quality into optimal,suboptimal,intermediate,and poor groups,with 759,329,133,and 125 instances,respectively.(2)The application of the Random Forest algorithm in gait quality assessment was explored.A relationship model was established between gait indicators and gait quality,yielding a predictive model accuracy of 95.99%.(3)The 13 main parameters significantly influencing asymmetric gait quality were identified through the Random Forest model's feature importance ranking.(4)An analysis of gait quality sensitivity factors using the 13 important parameters led to the identification of five key sensitivity indexes.The Random Forest model utilizing these sensitivity factors achieved a predictive accuracy of 94.20%.
ABSTRACT
Objective:To construct prediction models of necrotizing enterocolitis (NEC) using machine learning (ML) methods.Methods:From January 2015 to October 2021, neonates with suspected NEC symptoms receiving abdominal ultrasound examinations in our hospital were retrospectively analyzed. The neonates were assigned into NEC group (modified Bell's staging≥Ⅱ) and non-NEC group for diagnostic prediction analysis (dataset 1). The NEC group was subgrouped into surgical NEC group (staging≥Ⅲ) and conservative NEC group for severity analysis (dataset 2). Feature selection algorithms including extremely randomized trees, elastic net and recursive feature elimination were used to screen all variables. The diagnostic and severity prediction models for NEC were established using logistic regression, support vector machine (SVM), random forest, light gradient boosting machine and other ML methods. The performances of different models were evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, negative predictive value and positive predictive value.Results:A total of 536 neonates were enrolled, including 234 in the NEC group and 302 in the non-NEC group (dataset 1).70 were in the surgical NEC group and 164 in the conservative NEC group (dataset 2). The variables selected by extremely randomized trees showed the best predictive performance in two datasets. For diagnostic prediction models, the SVM model had the best predictive performance, with AUC of 0.932 (95% CI 0.891-0.973) and accuracy of 0.844 (95% CI 0.793-0.895). A total of 11 predictive variables were determined, including portal venous gas, intestinal dilation, neutrophil percentage and absolute monocyte count at the onset of illness. For NEC severity prediction models, the SVM model showed the best predictive performance, with AUC of 0.835 (95% CI 0.737-0.933) and accuracy of 0.787 (95% CI 0.703-0.871). A total of 25 predictive variables were identified, including age of onset, C-reactive protein and absolute neutrophil count at clincial onset. Conclusions:NEC prediction model established using feature selection algorithm and SVM classification model in ML is helpful for the diagnosis of NEC and grading of disease severity.
ABSTRACT
Herpes zoster is an infectious skin disease caused by reactivation of the varicella-zoster virus with multiple manifestations and various complication clinically. Studies have confirmed that chronic diseases are one of the independent risk factors for HZ; and the common chronic diseases such as diabetes, cardiovascular disease, chronic obstructive pulmonary disease and rheumatoid arthritis are associated with the development of herpes zoster. In this article, we review the latest research progress on the clinical features of herpes zoster, its correlation with common chronic diseases and the prevention strategies to reduce the disease burden.