Search

Recently, the rapid surge of reported COVID-19 cases attributed to the Omicron variant of severe acute respiratory syndrome coronavirus (SARS-CoV-2) created an immediate concern across nations. Local information pertaining to the new variant of concern (VOC) is lacking. We aimed to determine the clinical characteristics of COVID-19 during a period of Omicron prevalence among patients hospitalised from February 1 to 21, 2022 at Sungai Buloh Hospital and to estimate the risks of disease progression presumably caused by this variant in association with gender, age, comorbidity, and vaccination status.In this retrospective, singlecentered, retrospective cohort study, all hospitalised adults with laboratory-confirmed COVID-19, aged 18 and above, were recruited from February 1 to 21, 2022. Clinical characteristics, investigations, and outcomes were assessed.A total of 2279 patients aged 18 years and above with laboratory-proven COVID-19 were recruited and analysed, excluding 32 patients owing to incomplete data. Majority of the study population had a mean age of 41.8 ± 17.7, was female-predominant (1329/2279, 58.6%), had completed a primary series of vaccination with a booster (1103/2279, 48.4%), and had no underlying medical conditions (1529/2279, 67.4%). The risk of COVID-19-related disease progression was significantly lower in hospitalised patients under the age of 50 who were female, had no comorbidity, and had completed two doses of the primary series with or without a booster. [respectively, OR 7.94 (95% CI 6.16, 10.23); 1.68 (1.34, 2,12); 2.44 (1.85, 3.22); 2.56 (1.65, 3.97), p0.001].During the period of Omicron prevalence, a favourable outcome of COVID-19 was strongly associated with female gender, age below 50, a comorbidity-free condition, and having completed immunization. With this new observation, it could help improve public health planning and clinical management in response to the emergence of the latest VOC.

During the COVID-19 pandemic, bloodstream infection (BSI) rates were substantially rising in Sungai Buloh Hospital (HSB). It is believed that the COVID-19 pandemic has had an adverse impact on BSI incidence caused by contaminated periphery vascular catheters (PVCs). The study's objective is to reduce the BSI rates in HSB by improving adherence to the PVC care bundle via the Plan-Do-Study-Act (PDSA) approach.A quality improvement (QI) project was employed over four months, from June to September 2021, during the COVID-19 pandemic in HSB. All adults hospitalised for COVID-19 with intravenous lines were subjected to data collection. A baseline audit was conducted to study BSI incidence from April to May 2021. Implementation was carried out by PDSA cycles and data on BSI rates per 100 admissions was described using a monthly run chart.At baseline, the BSI rate per 100 admissions was 5.44 before implementing our QI project. Initial changes via PDSA cycles did not bring significant improvements to BSI rates and a rising trend in BSI rates was observed after two PDSA cycles. Further audits identified the problem of noncompliance with the practice of aseptic non-touch technique (ANTT) and a lack of effective leadership in implementing the PVC care bundle. The third PDSA cycle focused on adopting practical leadership skills among senior clinicians to ensure compliance with the prevention bundle and to encourage the use of ultrasound guidance for difficult line insertion. After the third PDSA cycle, the BSI rate per 100 admissions was reduced from 6.41 to 4.34 (p0.05). The BSI rates continued to decline down the line for another five months.Through QI initiatives, the risk of BSI can be significantly reduced.

In this study, machine learning (ML), a subdivision of artificial intelligence (AI), is implemented to study the mechanical behavior of adhesive single-lap joints (SLJs) subjected to tensile loading. The experimental data for training and testing the ML models are compiled from peer-reviewed journal papers to eliminate bias and increase the diversity within the data. The dataset is comprised of eight continuous SLJ parameters, which are used to predict the SLJ damage mode and failure strength. To accomplish this, regression and classification models are built using deep neural networks (DNN) and random forests (RF). Finite element (FE) modeling is conducted, and the performance is compared with the accuracy of the regression ML models. Results show ML models were able to predict strength with higher accuracy than FE modeling. Furthermore, both DNN and RF classified damage mode accurately without failure criteria, exposing limitations within FE modeling. As a result, this study introduces the use of ML for strength and damage mode prediction of adhesive SLJ, providing insights into their mechanical behavior, revealing hidden property performance patterns, and enhancing predictability.

The prominence of carbon fiber reinforced polymers (CFRP) has made advancements in many industries including aerospace, automotive, and sporting goods. Using additive manufacturing to produce the CFRPs allows prototyping complex geometries while keeping the structure lightweight and producing a high strength material. This work studies the printing parameters and flexural strength of 3D printed onyx based CFRP samples, looking at the effect of variable parameters affecting the carbon fiber layers. A Markforged Mark 2 3D printer, which utilizes a dual head print head, is used to print nine different configurations of the CFRP composite. Each sample consists of 24 layers of material with an infill of 50%, two wall layers, as well as four roof and floor layers. The configurations examine the extreme of three parameters with reference to the carbon fiber, being the fiber angle, number of fiber layers, and the position of the fiber layers. The carbon fiber print angle is printed with layers at 0˚, 0˚/45˚, and 0˚/90˚ with the latter two switching between the respective angles for every layer of carbon fiber printed. The number of fiber layers are set to be either 4 or 16. The position of fiber layers includes all layers (excluding roof/floor); the two outer most layers (which are directly below/above the roof/floor layers); and the four middle layers of the sample. The 3D printed CFRP specimens are subjected to three-point bend test, and three samples are tested for each configuration. One sample is chosen from each configuration to undergo pre-test and post-test inspection using X-ray micro-computed tomography to examine any internal damages.

About 40% of sea ice-covered areas have been reduced over the last three decades due to the effect of global warming. The Northern Sea route has been considered as a more effective, quicker, and economical sea route for marine vessels. But it is not safe to operate in such a cold and harsh environment due to the potential risk of collision with ice chunks, resulting in damages in the marine structures. It is therefore important to understand the behavior of marine composites at low temperatures, with the overarching goal that leads to improved design for marine structures and materials that can operate safely and effectively in the Arctic environment. This study investigates the low-velocity impact performance and damage mechanisms of woven carbon fiber reinforced polymer (CFRP), glass fiber reinforced polymer (GFRP), and carbon-glass fiber hybrid face sheets sandwich panel at both room temperature (23 °C) and low temperature (-70 °C) to mimic the Arctic environment. A series of low-velocity impact tests (3.46 m/s and 4.92 m/s) is performed at 15 J and 30 J energy using a drop tower testing machine. A relative comparison of the impact response due to different face sheet materials is presented in terms of force, displacement, and energy. Force-time and force-displacement plots show that GFRP sandwich composites have the highest damage initiation force and peak force values across different temperatures and impact energies. Furthermore, the lowest energy absorption for GFRP composites is found responsible for the least impact-induced damage. X-ray microcomputed tomography reveals severe fiber breakage on the compression side of CFRP sandwich panel and back face spitting at both temperatures for 30 J impacts. By replacing carbon fibers with glass fibers, the damage mechanism switches from fiber breakage to delamination as the dominant failure mode. The findings from this work will aid in a better understanding of the impact failure modes of composite sandwich structures at extremely low-temperature conditions.

With the development of poly (ADP-ribose) polymerase inhibitors for treatment of patients with cancer with an alteredIn this study, we built a new model (Asian Risk Calculator) for estimating the likelihood of carrying a pathogenic variant inAsian Risk Calculator included age of diagnosis, ethnicity, bilateral breast cancer, tumor biomarkers, and family history of breast cancer or ovarian cancer as predictors. The inclusion of tumor grade improved significantly the model performance. The full model was calibrated (Hosmer-LemeshowPopulation-specific customization of mutation prediction models and clinical genetic testing criteria improved the accuracy of BRCA mutation prediction in Asian patients.

Proxy Mobile IPv6 is an operational cluster of network-based localized mobility management protocol. This protocol has numerous advantages but still suffers from long latency and degradation in performance during the handover of video transmissions. In order to minimize the latency and improve performance during handover, this paper proposes a Signal to Interference and Noise Ratio (SINR)-based conversion and prediction approach. The approach sets up a prior connection based on the mobility prediction, reduces the time to acquire information, eliminates on-the-fly packet loss, and executes horizontal and vertical handover using SINR conversion. The motivation of this paper is to simulate, analyze and evaluate the video performance over Proxy Mobile IPv6 protocol. Simulations were conducted on packet and frame metrics which are latency, cumulative jitter, packet loss, frame loss and video quality measurements that analyzes and evaluates the network performances at the mobile host and during the handover.

Blockchain is a decentralized, distributed and public digital ledger technology. It can be visualized as a gradually increasing list of “blocks” which contains data that are linked together using cryptographic hash. Each transaction is verified by several participating nodes to compute a complex mathematical problem. The complexity of this computation, also known as Proof-of-Work (PoW), is governed by the difficulty set on a periodic basis. If the hash rate of the blockchain’s PoW grows or declines exponentially, the blockchain will be unable to maintain the block creation interval. The utilization of genetic algorithm (GA) in addition with the existing difficulty adjustment algorithm is proposed as a response to this by optimizing the blockchain parameters. A simulation of 3 scenarios as well as the default, were performed and the results were recorded. Based on the results, we are able to observe that the blockchain is able to reach the expected block time 74.4% faster than the blockchain without GA. Moreover, the standard deviations of the average block time and difficulty decreased by 99.4% and 99.5% respectively when block and difficulty intervals were considered for optimization, when compared to the default blockchain without GA.

A blockchain is a gradually increasing list of "blocks" containing information that are linked together using cryptographic hashes. It is a distributed, decentralized, and public digital ledger technology. Blockchains have caught not only the attention of researchers but also industries that are interested in implementing blockchains into existing products and services. However, performing studies on actual blockchain networks remain challenging as they may involve many nodes or these nodes could be placed in different geographical regions. Simulators for Proof-of-Work blockchain exists to facilitate this, but features such as difficulty adjustment or dynamic hash rate are not available. Using SimBlock, a blockchain network simulator as a base, difficulty adjustment algorithm and the capacity to increase or decrease hash rate dynamically were added to the simulator and the efficacy of the implementation was investigated for two scenarios; fixed hash rate and increasing hash rate periodically, over the length of the simulations. Based on the experiments, higher precision in terms of the average of actual time taken to mine a block were obtained, with a deviation of 0.68% compared to 5.73% in original SimBlock, and a deviation of 7.10% compared to 10.91% in original SimBlock, for fixed hash rate and dynamic hash rate adjustment respectively. Further studies were conducted for data from 2019 where the proposed implementation resulted in less than half the deviation achieved by the original implementation of SimBlock.

Tuberculosis is a nimble chameleon. It can manifest itself in various ways with atypical clinical and radiographic findings. In this report we discuss the importance of radiographic findings (nodular or mass-like forms) requiring a correlation with microbiological and histopathological results to differentiate lung cancer from TB.

There are many different protocols that regulate the way in which each node on a blockchain network is able to reach consensus that a newly created block is valid. One of the protocols, Proof-of-Work (PoW) gained popularity when it was implemented in a blockchain-based cryptocurrency known as Bitcoin. However, there are inherent deficiencies in its current implementation. This paper discusses these deficiencies, as well as the parameters that directly and indirectly affect its efficacy and performance so that possible enhancements to the protocol can be investigated.

Nowadays, low power operational amplifiers (op-amp) are highly demand for most of the applications such as in medical and communication system. In this project, two stage op-amp is designed and operated at 1.8 V supply voltage. The supply voltage is scaled down to reduce the power dissipation of the op-amp. This is because the power will be high when there is a large supply voltage. The design is simulated and analysed using Mentor Graphic Pyxis software. This two stage op-amp is designed using the Silterra 0.13 µm process technology. The operational amplifier provides a Direct Current (DC) gain of 21.18 dB and a unity gain bandwidth of 6.31 MHz. The gain margin obtained from the op-amp is 14.07 dB and the phase margin of the op-amp is 94.26 ° for 3 pF compensation capacitor and 10 pF load capacitor. The result shows that circuit able to work at 1.8 V power supply voltage and the total power dissipation for the op-amp is 5.35 mW.

Palm oil is the planet’s most exploited vegetable oil. However, its extensive commercialization has resulted in massive waste, particularly palm oil mill effluent (POME), contributing to severe environmental pollution. POME has a high concentration of oil and grease (O&G) with the mean value of 4,340 mg/L, exceeding the standard discharge limit of 50 mg/L. Hence, the recovery of oil content in POME is crucial as it could be a key material in biodiesel production. The oil droplets in POME exist in two phases: floating in the supernatant and suspended in the solids. During the solvent extraction process, the oil adsorbed by the solid particles is not entirely recovered. Thus, ultrasonication-based process intensification is introduced. Ultrasonication can break apart the solid particles and release the oil content using the principle of sound waves, thereby it will eventually increase the yield of oil recovery from POME. Although some studies were done on oil extraction from POME, the use of ultrasonication technique to enhance the extraction of oil from POME has never been done. The current research work is to investigate the feasibility of using ultrasonication technique to enhance the oil recovery from POME and compare it to a non-ultrasonicated POME. Overall, this research discovered that using ultrasonication as a pre-treatment would improve oil recovery yield from POME by 39.17% as compared to non-ultrasonicated sample under the optimum ultrasonication conditions of 30% amplitude and 30 seconds duration.

The Internet Engineering Task Force provides a network-based mobility management solution to execute handover in heterogeneous networks on network-side called Proxy Mobile IPv6 (PMIPv6). In this data article, data are presented during the horizontal and vertical handover on video communication in PMIPv6 mobility protocols. The handover data are gathered using several measurement factors, which are latency, jitter, cumulative measured, and peak signal noise ratio under network simulation software, for both horizontal and vertical handovers [8]. Keywords: Horizontal handover, Vertical handover, Video transmission, Wireless mobility, Seamless, Performance metrics, Data

Background Acral melanoma (AM) is the most common histopathological subtype of malignant melanoma in Asians. However, differences in the mutational profiles underlying AM and nonacral cutaneous melanoma (NAM) in Asians are not well understood. Objectives To augment the understanding of the prevalence, patterns and associations of various mutations between different subtypes of melanoma. Methods We performed comprehensive genomic profiling of 409 cancer-associated genes, using next-generation sequencing, in 66 primary melanomas comprised of 45 AMs and 21 NAMs. Results Most of the AMs (n = 27/45; 60%), but only five of 21 (24%) NAMs, were triple wild-type (triple-WT) tumours. Compared with AMs, NAMs exhibited a significantly higher frequency of BRAF mutations. The frequencies of NRAS/KRAS mutations, cell-cycle aberrations, copy number gains in BIRC2, BIRC3 and BIRC5, and gains of receptor tyrosine kinase genes were significantly higher in AMs. Ulceration was found at significantly higher rates in the AMs and NAMs with cell-cycle aberrations and gains of receptor tyrosine kinase genes. Notably, cell-cycle aberrations and copy number gains in BIRC2, BIRC3 and BIRC5 were significantly associated with poor melanoma-specific survival in the 66 patients with melanoma and especially in the 45 patients with AM. Multivariate analysis showed that lymph node metastasis and cell-cycle aberrations were independent prognostic factors of melanoma-specific survival. Conclusions This study strengthens our understanding of the patterns and clinical associations of oncogenic mutations in AMs and NAMs in Asians. What's already known about this topic? Mutation frequencies of driver genes vary between melanoma subtypes. Acral melanoma is the most common subtype of melanoma in Asians. KIT mutations and copy number variations occur more frequently in the acral subtype of melanoma than in the nonacral subtype What does this study add? NRAS/KRAS mutations, cell-cycle aberrations, copy number gains in BIRC2, BIRC3 and BIRC5, and amplifications of receptor tyrosine kinase genes were significantly enriched in acral melanoma and could be potential targets for treatment. Melanomas with cell-cycle aberrations and gains in receptor tyrosine kinase genes were significantly more likely to contain ulceration. What is the translational message? Cell-cycle aberrations and copy number gains in BIRC2, BIRC3 and BIRC5 were significantly associated with poor melanoma-specific survival. These observations should be explored further for future drug development.

The proliferation of video data has resulted in huge potential in Big Data technology and applications for video surveillance, security, multimedia tools and analytics. However, large-scale video data over a long period of time necessitates an efficient representation such that the task of retrieving video shots is rapid and accurate. This paper proposes an efficient and comprehensive framework for semantic based vehicle retrieval from long-term car park videos. Colour and motion semantics are respectively retrieved using intuitive colour term and sketch-based trajectory querying. The contribution of this work is twofold. First, we present a strategy for extracting the dominant colour for similarity matching against the Munroe ground truth tuples. Secondly, our proposed framework introduces a unique sketch-based method of retrieving vehicle motions, which relies on user-drawn trajectories. Using the spatio-temporal atom representation for extraction from videos, our approach obtained reasonably good precision scores at very fast retrieval speeds based on one-month long of daytime car park videos.

Traffic congestion is a significant problem for many developing countries. Among a variety of methods proposed to address this problem, extraction of road traffic conditions from microblogging platforms for dissemination to road users has gained popularity. Coupled with the use of Internet of Things (IoT) devices, up-to-date traffic condition reports from road users and road traffic authorities, can be effectively reconciled and communicated to the intended audience. However, the noisy and unstructured nature of Twitter texts causes a decline in the performance of conventional Named Entity Recognition (NER) techniques. NER techniques have traditionally been used to extract location information. In order to extract the corresponding traffic state information for each location, contextual information become important which are not implemented in conventional NER approaches. In this paper, a rule-based NER technique, which considers contextual clues, is proposed for the extraction of location and traffic state entities. The proposed approach has achieved significant improvement to the F1 score, 88.96 and 81.32, for location entity extraction from formal and informal sources, respectively.

With greater penetration of online services, the use of recommender systems to predict users’ propensity for continuous engagement becomes crucial in ensuring maximum revenue. There are many challenges, such as the cold start problem and data sparsity, that are continuously being addressed by a myriad of techniques in recommender systems. This paper provides insights into the trends of the techniques used for recommender systems and the challenges they address. With the insights; deep learning, matrix factorization or a combination of both can be used in addressing the data sparsity challenge.