Deep learning (DL) has been introduced in automatic heart-abnormality classification using ECG signals, while its application in practical medical procedures is limited. A systematic review is performed from perspectives of the ECG database, preprocessing, DL methodology, evaluation paradigm, performance metric, and code availability to identify research trends, challenges, and opportunities for DL-based ECG arrhythmia classification. Specifically, 368 studies meeting the eligibility criteria are included. A total of 223 (61%) studies use MIT-BIH Arrhythmia Database to design DL models. A total of 138 (38%) studies considered removing noise or artifacts in ECG signals, and 102 (28%) studies performed data augmentation to extend the minority arrhythmia categories. Convolutional neural networks are the dominant models (58.7%, 216) used in the reviewed studies while growing studies have integrated multiple DL structures in recent years. A total of 319 (86.7%) and 38 (10.3%) studies explicitly mention their evaluation paradigms, i.e., intra- and inter-patient paradigms, respectively, where notable performance degradation is observed in the inter-patient paradigm. Compared to the overall accuracy, the average F1 score, sensitivity, and precision are significantly lower in the selected studies. To implement the DL-based ECG classification in real clinical scenarios, leveraging diverse ECG databases, designing advanced denoising and data augmentation techniques, integrating novel DL models, and deeper investigation in the inter-patient paradigm could be future research opportunities.
Coronavirus disease 2019 (COVID-19) deaths can occur in hospitals or otherwise. In Malaysia, COVID-19 deaths occurring outside of the hospital and subsequently brought to the hospital are known as brought-in-dead (BID) cases. To date, the characteristics of BID COVID-19 cases in Malaysia are not clear. The objectives of this study are 2-fold: to explore the characteristics of 29,155 mortality cases in Malaysia and determine the factors associated with the high probability of BID, using the multilevel logistic regression model. Data on COVID-19 mortality cases from the entire country between March 17, 2020 and November 3, 2021 were retrieved from a national open data source. Of the 29,155 COVID-19 mortality cases, 5,903 (20.2%) were BID. A higher probability of BID (p < 0.05) was seen among individuals aged between 18 and 59 years, non-Malaysians, had no comorbidities, did not receive COVID-19 vaccination, and the interval between the date of death and diagnosis. A high prevalence of BID is an alarming public health issue, as this may signal health system failure at one or several levels and, hence, need urgent attention from relevant stakeholders. Based on the findings of this study, increasing the intensity of the vaccination campaign, addressing any issues faced by noncitizens about to COVID-19 management in- and out-of-hospital, increasing the awareness of signs and symptoms of worsening COVID-19 and, hence, the significance of self-monitoring, and determining the potential gaps in the health system may contribute to their increased risk of deaths.
Plasmodium knowlesi is an emerging species for malaria in Malaysia, particularly in East Malaysia. This infection contributes to almost half of all malaria cases and deaths in Malaysia and poses a challenge in eradicating malaria. The aim of this study was to develop a predictive model for P. knowlesi susceptibility areas in Sabah, Malaysia, using geospatial data and artificial neural networks (ANNs). Weekly malaria cases from 2013 to 2014 were used to identify the malaria hotspot areas. The association of malaria cases with environmental factors (elevation, water bodies, and population density, and satellite images providing rainfall, land surface temperature, and normalized difference vegetation indices) were statistically determined. The significant environmental factors were used as input for the ANN analysis to predict malaria cases. Finally, the malaria susceptibility index and zones were mapped out. The results suggested integrating geospatial data and ANNs to predict malaria cases, with overall correlation coefficient of 0.70 and overall accuracy of 91.04%. From the malaria susceptibility index and zoning analyses, it was found that areas located along the Crocker Range of Sabah and the East part of Sabah were highly susceptible to P. knowlesi infections. Following this analysis, targetted entomological mapping and malaria control programs can be initiated.
Background Older persons are at high-risk of developing severe complications from influenza. This consensus statement was developed to provide guidance on appropriate influenza prevention strategies relevant to the Malaysian healthcare setting. Methods Under the initiative of the Malaysian Influenza Working Group (MIWG), a panel comprising 11 multi-speciality physicians was convened to develop a consensus statement. Using a modified Delphi process, the panellists reviewed published evidence on various influenza management interventions and synthesised 10 recommendations for the prevention of influenza among the aged population via group discussions and a blinded rating exercise. Results Overall, annual influenza vaccination is recommended for individuals aged ≥ 60 years, particularly those with specific medical conditions or residing in aged care facilities (ACFs). There is no preference for a particular vaccine type in this target population. Antiviral agents can be given for post-exposure chemoprophylaxis or when vaccine contraindication exists. Infection control measures should serve as adjuncts to prevent the spread of influenza, especially during Hajj. Conclusion This consensus statement presents 10 evidence-based recommendations that can be adopted by healthcare providers to prevent influenza among the aged population in Malaysia. It could also serve as a basis for health policy planning in other lower- and middle-income countries.
For assessing the genetic diversity and genetic characterization of five Egyptian buffalo populations a total of 12 microsatellite markers were used. The total number of buffaloes sampled was 80, collected at random from five farms in five different governorates; Cairo, Kafr El-Sheikh, Shebeen El-Kom, Menoufia, and Beni Suef. The genetic parameters (allelic diversity, allelic frequencies, observed heterozygosity, unbiased expected heterozygosity, and polymorphic information content) were calculated using three different programs. All used microsatellites were polymorphic and ranged from four alleles (Loci; CSSM029, CSSM036, CSSM038, CSSM043, CSSM046, and ILSTS005) to nine alleles (Loci; BM1818 and CSSM047) with a total of 64 alleles in the whole population. Allelic richness for the whole population ranged between 3.297 (in locus CSSM029) and 6.806 (in locus CSSM047) with overall mean 4.574. Within populations, Kafr El-Sheikh population had the highest average of allelic richness (4.384). This indicates the potential of this population to adapt with environmental changes in future compared with other populations. BMC1013, BM1818, CSSM019, and CSSM047 showed the highest allelic richness. PIC estimates were high and ranged between 0.65 (in locus CSSM029) and 0.92 (in locus CSSM022) with an average of 0.82. Values of H o were lower than values of H Nb for all populations, which denoting depression of heterozygotes in these populations and may be attributable to existence of null alleles and inbreeding. This study as well proves the usefulness of heterologous bovine microsatellite markers in evaluation of the genetic variability in Egyptian buffalo populations due to high polymorphism, informativeness of these markers which can be used to develop future breeding strategies and conservation decisions on our indigenous breed.
Understanding COVID 19 cluster infection is vital as it evaluates the current situation and serves as the basis of further action in control and prevention strategies. We aim to describe the characteristics of COVID-19 clusters in Malaysia based on location, types, positive percentage, and case fatality ratio (CFR). We used open-source data of COVID-19 clusters from the GitHub Ministry of Health Malaysia website. The data were downloaded, cleaned, and analysed using SPSS version 27. The analysis includes data of clusters that have been declared as ended from 1st March 2020 to 10th August 2021. A total of 3,495 clusters of COVID19 were reported in Malaysia involved 317,935 confirmed cases, representing 24.4% of total cases in the country within the same period. The majority of the clusters occurred in a single state (88.1%) compared to multiple states' involvements. There were increasing trends of reporting clusters and more involvement in workplace and community clusters. Workplace clusters represent the highest percentage of all clusters (54.1%). The positive percentage of COVID-19 testing was highest with a detention centre cluster (32.9%); meanwhile, CFR was highest in the cluster of high-risk populations. Strategic action in controlling and preventing COVID-19 has to be focused on high-risk areas such as the workplace. More COVID-19 screening should be done in clusters involving high-risk populations and institutions such as detention centres.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.