Asthma Classification Research Articles

Machine Learning-Based Risk Prediction of Asthma

This paper will attempt to detect asthma early based on machine learning. This work utilizes a dataset consisting of 21 features such as family history of asthma, intake of drugs, smoking status, and quality of air and much more. The work tries to develop models with maximum precision, recall, and accuracy in asthma classification. These include several machine learning algorithms evaluated: SVM, Decision Trees, and Random Forest. This study focuses on both medical and environmental data when used in combination. The results here show that the Random Forest model performed well in achieving an overall high performance, with an accuracy of 0.926829. This has dire implications for evaluation and early diagnosis of Asthma. Key Words: Asthma detection, Machine learning, Patient history, Environmental factors, Support Vector Machines (SVM), Decision Trees, Random Forests, Predictive models, Air quality, Smoking habits, Medication usage, Family asthma history, Ac- curacy, Precision, Recall, Early diagnosis, Healthcare, Medical data, Environmental data, Predictive power

Evaluating Severe Therapy-Resistant Asthma in Children: Diagnostic and Therapeutic Strategies

Introduction: Worldwide, asthma is the most common non-communicable respiratory disease and causes considerable morbidity and mortality. Most people with asthma can be treated effectively with low-dose medications if these are taken correctly and regularly. Around 10% of people with asthma have an uncontrolled form of the disease or can only achieve control with high-dose medications, incurring disproportionately high health care costs. Areas Covered: PubMed and personal archives were searched for relevant articles on the definition, management and pharmacotherapy of severe asthma. The WHO classification of severe asthma and the treatment levels encompassed in the definition are discussed. Most children and young people referred for consideration of ‘beyond-guidelines therapy’ can in fact be managed on standard treatment after a multi-disciplinary team assessment focusing on ensuring correct basic management, and these steps are described in detail. Options for those with true therapy-resistant asthma are described. These include monoclonal antibodies, most of which target type 2 inflammation. Expert Opinion: Getting the basics right is still the most important aspect of asthma care. For those with severe, therapy-resistant asthma, an increasing number of life-transforming monoclonals have been developed, but there is still little understanding of, and a paucity of treatment options for, non-eosinophilic asthma.

Bayesian additive regression trees for predicting childhood asthma in the CHILD cohort study

BackgroundAsthma is a heterogeneous disease that affects millions of children and adults. There is a lack of objective gold standard diagnosis that spans the ages; instead, diagnoses are made by clinician assessment based on a cluster of signs, symptoms and objective tests dependent on age. Yet, there is a clear morbidity associated with chronic asthma symptoms. Machine learning has become a popular tool to improve asthma diagnosis and classification. There is a paucity of literature on the use of Bayesian machine learning algorithms to predict asthma diagnosis in children. This paper develops a prediction model using the Bayesian additive regression trees (BART) and compares its performance to various machine learning algorithms in predicting the diagnosis of childhood asthma.MethodsClinically relevant variables collected at or before 3 years of age from 2794 participants in the CHILD Cohort Study were used to predict physician-diagnosed asthma at age 5. BART and six other commonly used machine learning algorithms, namely adaptive boosting, logistic regression, decision tree, neural network, random forest, and support vector machine were trained. Measures of performance including sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve were calculated. The confidence intervals were calculated using Bootstrapping samples. Important predictors and interaction effects associated with asthma were also identified using BART.ResultsBART, logistic regression and random forest showed the highest area under the ROC curve compared to other machine learning algorithms. Based on BART, recurrent wheeze, respiratory infection and food sensitization at 3 years of age were the most important predictors. The three most important interaction effects were found to be interaction terms of respiratory infection at 3 years and recurrent wheezing at 3 years, maternal asthma and paternal asthma, and maternal wheezing and inhalant sensitization of child at 3 years.ConclusionsBART demonstrated promising prediction performance when compared to other machine learning algorithms. Future research could validate the BART in an external cohort to evaluate its reliability and generalizability.

Metabolic and ionic control of T cells in asthma endotypes.

CD4+T cells play a central role in orchestrating the immune response in asthma, with dysregulated ion channel profiles and altered metabolic signatures contributing to disease progression and severity. An important classification of asthma is based on the presence of T-helper cell type 2 (Th2) inflammation, dividing patients into Th2-high and Th2-low endotypes. These distinct endotypes have implications for disease severity, treatment response, and prognosis. By elucidating how ion channels and the energy metabolism control Th cells in asthma, this review contributes to the pathophysiological understanding and the prospective development of personalized therapeutic treatment strategies for patients suffering from distinct asthma endotypes.

Real-world phenotyping and risk assessment of childhood asthma burden using national registries

BackgroundPhenotype classification contributes to risk assessment of asthma. Previous studies have applied this concept primarily to adult populations and in the setting of research protocol assessments which may not be applicable to clinical settings. ObjectiveExploring the value of routinely collected clinical data for phenotype classification and risk assessment of childhood asthma. MethodsUsing hospital and laboratory data, 29,851 children in a Danish nationwide database aged 2–17 years with ICS-treated asthma in 2015 followed for two years (730 days) were classified to have T2 (elevated blood eosinophils (>300 cells/μL) and/or elevated total- or specific-IgE), and/or non-T2 risk factors (in utero tobacco exposure and/or severe viral infections). Logistic regression was applied to quantify associations of risk factors with asthma severity, control, and exacerbation risk. ResultsIn a complete case analysis, 85.8 % children had at least one T2 risk factor and 29.3 % had mixed T2/non-T2 risk factors. Elevated blood eosinophils and total/specific IgE were associated with exacerbations (ORs 1.55 (1.38–1.73) and 1.41 (1.20–1.66) and higher asthma severity (1.42 (1.24–1.63) and 1.31 (1.08–1.60)), respectively.Dose-dependency was observed between blood eosinophil counts, total IgE levels, and risk of adverse outcomes. Furthermore, accumulation of risk factors demonstrated an increasing risk, with children with all four risk factors having a high risk of any adverse asthma-related outcome (OR 3.13 (2.03–4.82) ConclusionAsthma phenotypic markers defined in research protocols can be reliably applied in real-world settings by utilizing data collected during routine clinical care and enable better classification of risk of adverse asthma outcomes.

Small airway inflammation in atypical asthma

BackgroundLess attention has been paid to the pathophysiological changes in atypical asthma such as cough variant asthma (CVA) and chest tightness variant asthma (CTVA). The obstruction of large and small airways is the important component in the development of asthma. We investigated whether small airway inflammation (SAI) induced small airway dysfunction (SAD) in these atypical asthmatics. MethodsSix hundred and eighty-six patients were enrolled and analyzed in the study. The partitioned airway inflammation was assessed by fractional exhaled nitric oxide (FeNO), such as FnNO, FeNO50, FeNO200, and calculated alveolar fraction of exhaled NO (CaNOdual). Correlations between exhaled NOs and SAD-related variables were assessed, whereas cell classification was evaluated by Spearman's rank tests. Classic asthma, CVA, and CTVA about potential risk were conducted using binary logistic regression models. ResultsThe whole airway inflammation increased in classic and atypical asthma than controls, whereas the central and peripheral airway inflammation in the CVA and CTVA groups increased compared with the classic asthma group. Smoking exposure was found to increase the central and peripheral airway inflammation in patients with asthma. The exhaled NO of FeNO50 and FeNO200 was associated with SAD in classic asthma, but not in CVA and CTVA. FeNO200 was the main risk (adjusted odds ratio [OR], 1.591; 95 % CI, 1.121–2.259; P = .009) in classic asthma and (adjusted OR, 1.456; 95 % CI, 1.247–1.700; P = .000) in CVA. The blood eosinophil levels were correlated with FeNO50 and FeNO200 in classic asthma and atypical asthma. ConclusionMore severe inflammatory process was present in central and peripheral airways in CVA and CTVA, which might reflect a pre-asthmatic state. SAI was the predominant risk factor in the development of asthma before SAD.

The burden of allergic rhinitis is undermanaged in a large proportion of Chinese young adults from Singapore

BackgroundAllergic rhinitis (AR) is a nasal disorder characterized by the simultaneous manifestation of at least 2 out of 4 possible symptoms: rhinorrhea, nasal itching, nasal pruritus, and sneezing. Presently, among Chinese young adults from Singapore, we characterised AR phenotypes, established Total Nasal Symptom Score (TNSS) baselines, and examined the management of AR. MethodsParticipants completed an investigator-administered International Study of Asthma and Allergies in Childhood (ISAAC) questionnaire and underwent a skin prick test (SPT). Individuals exhibiting sensitization during the SPT while having at least 2 rhinitis symptoms were identified as AR cases, then categorized into Allergic Rhinitis in Asthma (ARIA) classifications. ResultsThere were 9323 subjects analyzed. AR prevalence was estimated at 35.4%. Rhinorrhea was perceived as the most severe (mean Nasal Symptom Score (mNSS) ± SD: 1.42 ± 0.74), while nasal pruritus was the least severe (mNSS ± SD: 1.24 ± 0.68). Among moderate-severe AR (68.1%), most were affected by either troublesome symptoms (27.7%) or sleep disturbances (18.4%). By ARIA classes, 26.6% were mild intermittent, 5.4% were mild persistent, 50.3% were moderate-severe intermittent, and 17.6% were moderate-severe persistent. The mean TNSS (mTNSS) of AR cases was 4.43 (SD = 2.49) and between AR classifications, the mTNSS was significantly different. Notably, a large proportion of AR cases remained undiagnosed (85.2%), untreated (72.5%), or both (65.4%); 19.8% self-medicated for AR. ConclusionsThere was a significant difference in TNSS of the AR phenotypes, and among phenotypes with a higher mTNSS, a large proportion remained untreated, undiagnosed, or both. The evidence indicates an existing burden of AR among Chinese young adults in Singapore which is notably undermanaged.

Non-Invasive Detection of Cytokines in Saliva Towards Asthma Phenotyping Using an Electrochemical Sensor

Introduction: Asthma is a complicated and heterogeneous condition of the lungs which is usually associated with the chronic inflammation of the airways. Traditionally, classification of asthma has been based on clinical attributes such as the symptoms observed and response to treatment. However, with the growing understanding of the heterogeneity of asthma as a spectrum with varying underlying causes, asthma phenotyping has been on the horizon. Cytokines are the chemical messengers that play a crucial role in the immune system. They mediate communication between the cells in the immune system and as such have a direct effect on the inflammation process. The role of some cytokines has been established in the pathogenesis of asthma. Due to the involvement of cytokines in the modulation of asthma, they can be excellent target for asthma phenotyping. This is also helpful because identification and quantification of these inflammatory mediators can allow for more precise medicine for more targeted treatment. To aid the identification and quantification of cytokines in the phenotyping of asthma, we have created a customizable (currently for IL-8 and IL-10) multi-marker electrochemical platform for the detection of cytokines non-invasively in saliva. The developed electrochemical platform can detect cytokines rapidly and non-invasively within reported physiological ranges of the given cytokines. Evaluating the presence and quantity of the cytokines will be useful in the phenotyping and treatment plans adopted by physicians.Materials and Methods: The electrochemical sensor consists of 6 multiplexed gold electrodes on a PCB substrate. IL-8 and IL-10 antibodies were immobilized onto the sensor surface using a gold-thiol linker chemistry. Sensor response was evaluated with non- faradaic Electrochemical Impedance Spectroscopy (nf-EIS). Pooled human saliva was spiked with increasing doses of IL-8 and IL-10 in their respective physiological ranges to develop a dose response curve. Spike and recovery experiments were conducted to validate the efficacy of dose response curve.Results and Discussion: Sensor response to both IL-8 and IL-10 spiked doses in pooled saliva showed a dose-dependent response to increasing concentrations of IL-8 and IL-10 respectively as illustrated by the Nyquist plots in Fig.1. The change in impedance with increasing doses of IL-8 and IL-10 in pooled human saliva lends to credence to the establishment of the presence and quantification of cytokines in saliva that can be used by clinicians and scientists alike in the further development of personalized medicine for severe asthmatics unresponsive to traditional asthmatic treatments.Conclusion: This study presents a noninvasive electrochemical sensor for rapid and early detection of cytokines in pooled human saliva towards asthma phenotyping. The sensor exhibits a dose dependent response to IL-8 and IL-10, markers implicated in pathogenesis of asthma. Sensor capability was demonstrated with spike and recovery experiments illustrating the ability of the developed electrochemical sensor to detect and quantify cytokines in human saliva. Figure 1

Fourier Decomposition-Based Automated Classification of Healthy, COPD, and Asthma Using Single-Channel Lung Sounds

Fourier Decomposition-Based Automated Classification of Healthy, COPD, and Asthma Using Single-Channel Lung Sounds

Endotypes of severe neutrophilic and eosinophilic asthma from multi-omics integration of U-BIOPRED sputum samples.

Clustering approaches using single omics platforms are increasingly used to characterise molecular phenotypes of eosinophilic and neutrophilic asthma. Effective integration of multi-omics platforms should lead towards greater refinement of asthma endotypes across molecular dimensions and indicate key targets for intervention or biomarker development. To determine whether multi-omics integration of sputum leads to improved granularity of the molecular classification of severe asthma. We analyzed six -omics data blocks-microarray transcriptomics, gene set variation analysis of microarray transcriptomics, SomaSCAN proteomics assay, shotgun proteomics, 16S microbiome sequencing, and shotgun metagenomic sequencing-from induced sputum samples of 57 severe asthma patients, 15 mild-moderate asthma patients, and 13 healthy volunteers in the U-BIOPRED European cohort. We used Monti consensus clustering algorithm for aggregation of clustering results and Similarity Network Fusion to integrate the 6 multi-omics datasets of the 72 asthmatics. Five stable omics-associated clusters were identified (OACs). OAC1 had the best lung function with the least number of severe asthmatics with sputum paucigranulocytic inflammation. OAC5 also had fewer severe asthma patients but the highest incidence of atopy and allergic rhinitis, with paucigranulocytic inflammation. OAC3 comprised only severe asthmatics with the highest sputum eosinophilia. OAC2 had the highest sputum neutrophilia followed by OAC4 with both clusters consisting of mostly severe asthma but with more ex/current smokers in OAC4. Compared to OAC4, there was higher incidence of nasal polyps, allergic rhinitis, and eczema in OAC2. OAC2 had microbial dysbiosis with abundant Moraxella catarrhalis and Haemophilus influenzae. OAC4 was associated with pathways linked to IL-22 cytokine activation, with the prediction of therapeutic response to anti-IL22 antibody therapy. Multi-omics analysis of sputum in asthma has defined with greater granularity the asthma endotypes linked to neutrophilic and eosinophilic inflammation. Modelling diverse types of high-dimensional interactions will contribute to a more comprehensive understanding of complex endotypes. Unsupervised clustering on sputum multi-omics of asthma subjects identified 3 out of 5 clusters with predominantly severe asthma. One severe asthma cluster was linked to type 2 inflammation and sputum eosinophilia while the other 2 clusters to sputum neutrophilia. One severe neutrophilic asthma cluster was linked to Moraxella catarrhalis and to a lesser extent Haemophilus influenzae while the second cluster to activation of IL-22.

Occupational bronchial asthma: a clinician’s view

The author provides an overview of the etiological factors and classification of occupational bronchial asthma and describes the coding of occupational asthma according to ICD 10, the pathogenesis of the disease, the clinical picture, and the diagnostic procedure. The work presents contraindications for admission to work and the procedure for conducting a medical and social examination; the recommendations for patient rehabilitation are provided.

Novel Machine Learning Identifies 5 Asthma Phenotypes Using Cluster Analysis of Real-World Data

Asthma classification into different subphenotypes is important to guide personalized therapy and improve outcomes. To further explore asthma heterogeneity through determination of multiple patient groups by using novel machine learning (ML) approaches and large-scale real-world data. We used electronic health records of patients with asthma followed at the Cleveland Clinic between 2010 and 2021. We used k-prototype unsupervised ML to develop a clustering model where predictors were age, sex, race, body mass index, prebronchodilator and postbronchodilator spirometry measurements, and the usage of inhaled/systemic steroids. We applied elbow and silhouette plots to select the optimal number of clusters. These clusters were then evaluated through LightGBM's supervised ML approach on their cross-validated F1 score to support their distinctiveness. Data from 13,498 patients with asthma with available postbronchodilator spirometry measurements were extracted to identify 5 stable clusters. Cluster 1 included a young nonsevere asthma population with normal lung function and higher frequency of acute exacerbation (0.8 /patient-year). Cluster 2 had the highest body mass index (mean ± SD, 44.44 ± 7.83 kg/m2), and the highest proportion of females (77.5%) and Blacks (28.9%). Cluster 3 comprised patients with normal lung function. Cluster 4 included patients with lower percent of predicted FEV1 of 77.03 (12.79) and poor response to bronchodilators. Cluster 5 had the lowest percent of predicted FEV1 of 68.08 (15.02), the highest postbronchodilator reversibility, and the highest proportion of severe asthma (44.9%) and blood eosinophilia (>300 cells/μL) (34.8%). Using real-world data and unsupervised ML, we classified asthma into 5 clinically important subphenotypes where group-specific asthma treatment and management strategies can be designed and deployed.

Economic burden assessment for the management of asthma patients at Mexico's National Institute for Respiratory Diseases

Determinar la carga económica anual del asma, desde una perspectiva institucional y con base en la clasificación recomendada por GINA, en una cohorte retrospectiva de adultos atendidos en el Instituto Nacional de Enfermedades Respiratorias (INER) de México. Estudio observacional, longitudinal y retrospectivo, llevado a cabo a partir de la información recabada de 247 pacientes femeninas con asma. Se estimaron los costos directos anuales: visitas, pruebas de laboratorio, tratamiento farmacológico y de las crisis o exacerbaciones, para determinar la carga anual de la enfermedad desde una perspectiva institucional, y según la clasificación de la Iniciativa Global para el Asma. El costo promedio anual fue de $43,813,92, que aumentó en relación con la necesidad de aumento de dosis de corticoides inhalados y beta-agonistas de acción prolongada. El costo promedio de la consulta médica fue de $2004.57, $982.82 por gestión de crisis y $2645.95 por pruebas de laboratorio. El tratamiento farmacológico representó la principal carga económica, con un costo promedio anual de $38,180.58. Los resultados resaltan una carga económica del asma estimada en un costo anual por paciente de $43,813.92 MXN (DE=93,348.85), en el contexto del tercer nivel de atención en el sistema de salud público mexicano. La gravedad del asma, los tratamientos y los biológicos fueron los principales factores que aumentaron los costos directos de la atención.

Insufficiency of Laboratory Data in Reflecting Allergic Rhinitis Severity Based on the Allergic Rhinitis and Its Impact on Asthma Guideline in Korean Patients

Background and Objectives: This retrospective study, conducted at a single tertiary medical center, aimed to investigate the correlation between the severity of allergic rhinitis (AR) based on subjective symptoms and the severity assessed through laboratory data.Methods: In total, 584 patients who were diagnosed with AR by a multiple-allergen simultaneous test were included. Patients were classified into four groups according to the Allergic Rhinitis and its Impact on Asthma (ARIA) classification guideline. The visual analog scale (VAS) score for overall discomfort and laboratory parameters, including the serum total immunoglobulin E (IgE) level and peripheral blood eosinophil count, were evaluated in all patients. An analysis was conducted to examine the differences in VAS scores and laboratory findings among the four groups. Additionally, the correlations between the laboratory findings and VAS score were analyzed.Results: The serum total IgE level and the percentage and count of peripheral blood eosinophils showed no significant differences among the groups. However, the VAS score for overall discomfort exhibited notable between-group differences. The average VAS score was 6.14 (95% confidence interval 5.94–6.34) in the overall group. The mean scores of each group showed a noticeable increasing trend from the mild intermittent group to the mild persistent, moderate to severe intermittent, and moderate to severe persistent groups (p<0.001), although there was no clear correlation between the increase in VAS scores and laboratory parameters.Conclusion: Neither the symptom-based ARIA guideline nor the VAS score correlated with the AR laboratory test measurements. The current laboratory data alone may not be sufficient to reflect the severity of AR based on subjective symptoms.

In asthma patients: relationship between symptom control, quality of life, and the beck anxiety scale

Aims: Asthma is a chronic disease characterized by recurrent airway obstruction. It affects the quality of life of patients and can lead to anxiety and other psychological disorders. In this study, we aimed to find out how asthma affects patients' quality of life and psychological state by using the SF-36 Quality of Life Scale and the Beck Anxiety Scale (BAS) in patients diagnosed with asthma. Methods: 222 patients diagnosed with asthma in our outpatient clinics were included in the study. Patient demographic characteristics and respiratory function test parameters were recorded. Outcomes were determined by assessing asthma control status with the Asthma Control Test (ACT), patients' emotional state with the BAS, and quality of life with the SF-36. Results: According to ACT, 37.8% of patients were uncontrolled, 33.8% were partially controlled, and controlled 28.4%. The distribution in terms of asthma classification: 29.8% were mild intermittent, 5% mild persistent, 35.1% moderate persistent and 40.1% severe persistent. When comparing between ACT groups in relation to the SF-36 quality of life scale; physical function (p<0.001), physical role difficulties (p<0.001), emotional role difficulties (p<0.001), vitality (p<0.001), mental health (p<0.001), social function (p<0.001), pain (p<0.001), and general health (p<0.001). There was a difference between the groups ACT in relation to BAS (p<0.001). Conclusion: We believe that when regulating the medical treatment of asthma patients, not only their respiratory functions but also their quality of life and psychological status should be evaluated, and measures should be taken to improve the quality of life and psychological disorders.

Asthma is associated with increased severity and duration of rhinitis: A study with the Allergic Rhinitis and its Impact on Asthma classes in the Constances cohort.

Few population-based studies have described allergic rhinitis (AR) according to the Allergic Rhinitis and its Impact on Asthma (ARIA) classification, and none have assessed the impact of asthma on this classification. Our aims were to 1) describe AR according to four ARIA classes and 2) within each of the four ARIA classes, compare participants with AR alone versus those with AR and asthma. Cross-sectional analyses were performed using data from the 2014 annual follow-up questionnaire of the French adult population-based cohort Constances. Current AR was defined by the report of sneezing, runny, or blocked nose in the last 12months and the report of nasal allergies. Following ARIA recommendations, rhinitis was classified according to its severity (mild or moderate-severe) and duration (intermittent or persistent). Ever asthma was also defined by a questionnaire. Among the 4675 participants with AR (57% women, mean age 50.2±12.7years), 44% were classified as mild/intermittent, 16% mild/persistent, 25% moderate-severe/intermittent, and 15% moderate-severe/persistent. Within each of the four ARIA classes, compared to participants with rhinitis alone, participants with rhinitis and asthma had significantly more severe symptoms, more conjunctivitis, a higher mean eosinophil count and more treatments with intra-nasal corticosteroids and oral antihistamines co-medication. This is a paradigm shift study as for the first time this large population-based study in adults showed that asthma status has a profound effect on the ARIA classification. Rhinitis alone and rhinitis with asthma represent two distinct phenotypes. These results reinforce the need to include asthma status in the ARIA classification and guidelines.

Biomarkers associated with severity classification of asthma identified by comprehensive bioinformatics analysis

Using an integrated bioinformatics approach to find novel biomarkers that can predict asthma severity. From June 2022 to December 2022, this clinical medical study was conducted and completed in the Department of Allergy, Zhongnan Hospital of Wuhan University. The gene chip dataset GSE43696 was screened and downloaded from the high-throughput Gene Expression Omnibus (GEO) database, and the gene chip data preprocessing was completed using package "affy" in R and "rma" algorithm in turn. Use the the "edgeR" and "limma" packages to screen out the differentially expressed genes (DEGs) between normal controls, mild to moderate asthma patients and severe asthma patients, and then use the "clusterProfiler" package to perform GO enrichment analysis and KEGG pathway enrichment analysis of DEGs, finally use the STRING website to construct a protein-protein interaction (PPI) network of DEGs to further screen key genes. Using the R language "WGCNA" package, the weighted gene co-expression network analysis (WGCNA) was performed on the dataset GSE43696, and the modules significantly related to the severity of asthma were screened out, then the hub genes were obtained by intersecting the WGCNA analysis results with the DEGs screened by PPI. Datasets GSE43696 and GSE63142 were used to verify the expression of hub genes, and the diagnostic value was evaluated according to the ROC curve, then the potential function of hub genes in dataset GSE43696 was further clarified by gene set enrichment analysis (GSEA). The results showed that a total of 251 DEGs were screened, including 39 in the normal group and mild to moderate asthma group, 178 in the normal group and severe asthma group, and 34 in the mild to moderate asthma group and severe asthma group, mainly involved in biological processes such as response to toxic substance, response to oxidative stress, extracellular structure organization, extracellular matrix organization. Two modules significantly correlated with asthma severity were screened out (red module, P=7e-6, r=0.43; pink module, P=5e-8, r=-0.51), and finally six hub genes were obtained, including B3GNT6, CEACAM5, CCK, ERBB2, CSH1 and DPPA5. The comparison of gene expression levels and ROC curve analysis of datasets GSE43696 and GSE63142 further verified the six hub genes, which may associated with o-glycan biosynthesis, alpha linolenic acid metabolism, linoleic acid metabolism, pentose and glucoronate interconversions. In conclusion, through a variety of bioinformatics analysis methods, this study identified six hub genes significantly related to the severity of asthma, which potentially provided a new direction for the prediction and targeted therapy of asthma.

RDN-NET: A Deep Learning Framework for Asthma Prediction and Classification Using Recurrent Deep Neural Network

Asthma is the one of the crucial types of disease, which causes the huge deaths of all age groups around the world. So, early detection and prevention of asthma disease can save numerous lives and are also helpful to the medical field. But the conventional machine learning methods have failed to detect the asthma from the speech signals and resulted in low accuracy. Thus, this paper presented the advanced deep learning-based asthma prediction and classification using recurrent deep neural network (RDN-Net). Initially, speech signals are preprocessed by using minimum mean-square-error short-time spectral amplitude (MMSE-STSA) method, which is used to remove the noises and enhances the speech properties. Then, improved Ripplet-II Transform (IR2T) is used to extract disease-dependent and disease-specific features. Then, modified gray wolf optimization (MGWO)-based bio-optimization approach is used to select the optimal features by hunting process. Finally, RDN-Net is used to predict the asthma disease present from speech signal and classifies the type as either wheeze, crackle or normal. The simulations are carried out on real-time COSWARA dataset and the proposed method resulted in better performance for all metrics as compared to the state-of-the-art approaches.

Development and validation of asthma risk prediction models using co-expression gene modules and machine learning methods

Asthma is a heterogeneous respiratory disease characterized by airway inflammation and obstruction. Despite recent advances, the genetic regulation of asthma pathogenesis is still largely unknown. Gene expression profiling techniques are well suited to study complex diseases including asthma. In this study, differentially expressed genes (DEGs) followed by weighted gene co-expression network analysis (WGCNA) and machine learning techniques using dataset generated from airway epithelial cells (AECs) and nasal epithelial cells (NECs) were used to identify candidate genes and pathways and to develop asthma classification and predictive models. The models were validated using bronchial epithelial cells (BECs), airway smooth muscle (ASM) and whole blood (WB) datasets. DEG and WGCNA followed by least absolute shrinkage and selection operator (LASSO) method identified 30 and 34 gene signatures and these gene signatures with support vector machine (SVM) discriminated asthmatic subjects from controls in AECs (Area under the curve: AUC = 1) and NECs (AUC = 1), respectively. We further validated AECs derived gene-signature in BECs (AUC = 0.72), ASM (AUC = 0.74) and WB (AUC = 0.66). Similarly, NECs derived gene-signature were validated in BECs (AUC = 0.75), ASM (AUC = 0.82) and WB (AUC = 0.69). Both AECs and NECs based gene-signatures showed a strong diagnostic performance with high sensitivity and specificity. Functional annotation of gene-signatures from AECs and NECs were enriched in pathways associated with IL-13, PI3K/AKT and apoptosis signaling. Several asthma related genes were prioritized including SERPINB2 and CTSC genes, which showed functional relevance in multiple tissue/cell types and related to asthma pathogenesis. Taken together, epithelium gene signature-based model could serve as robust surrogate model for hard-to-get tissues including BECs to improve the molecular etiology of asthma.

1-Year Prospective Study of the Relationship of Serial Exhaled Nitric Oxide Level and Asthma Control.

Previous studies found that the fractional nitric oxide concentration in exhaled breath (FeNO) levels in healthy Chinese adults was higher than in White adults. More understanding of serial changes of FeNO levels with asthma control in a real-life clinical setting would be important to explore the utility of this biomarker in routine asthma management. This study assessed the FeNO levels of Chinese asthma subjects with different levels of asthma control and the serial changes with respect to the changes in asthma control over 1 year. A 12-month prospective study (subjects recruited between November 2019 and January 2021) with serial measurement of FeNO levels at baseline, 4, 8 and 12 months. Asthma control was assessed by the Global Initiative for Asthma classification, Asthma Control Test (ACT) and Asthma Control Questionnaire (ACQ). Altogether, 136 subjects (mean age 51.51±15.09 years, 46[33.8%] male) had successful baseline FeNO measurements. At baseline, the FeNO levels did not show a statistically significant difference for controlled, partly controlled and uncontrolled asthma according to GINA classification, ACT and ACQ. FeNO levels decreased with improving asthma control and stayed at similar levels with unchanged or worsening asthma control for all subjects. For subjects with baseline blood eosinophil levels ≥300 cells/µL(n=59), FeNO levels decreased with improving asthma control, stayed similar without change for asthma control and increased with worsening asthma control. Receiver operating characteristic (ROC) analysis with the highest area under curve (AUC) for changes in FeNO levels for improving asthma control was between ≤ -10 to -25 ppb at various time points in the 12-month study. Changes in FeNO levels over time were associated with changes in clinical asthma control, particularly in those with higher blood eosinophil count and are likely more useful than a single time point measurement in managing asthma.