Role of Interfacial Tension on Viscous Multiphase Flows in Coaxial Microfluidic Channels.

We experimentally investigate the influence of interfacial tension on liquid/liquid microflows for fluids having large viscosity contrasts. A coaxial microdevice is employed to examine the situation where a less-viscous fluid is injected in a sheath of a more-viscous fluid using both immiscible and miscible fluid pairs. Data obtained from high-speed imaging reveal a variety of regular flow regimes, including dripping, jetting, wavy, core-annular, diffusive jet, mist, and inverted thread flow patterns. Flow maps are delineated over a wide range of injection flow rates, and an original methodology based on periodic pattern analysis is developed to clarify relationships between interfacial dynamics and fluid properties of multiphase materials. Specifically, we show the smooth evolution of droplet size and spacing at the transition between dripping and jetting flows and develop scaling relationships based on capillary numbers to predict droplet flow morphologies. For similar flow conditions, reducing interfacial tension leads to a significant decrease in droplet size. For miscible fluid pairs, diffusive jets are observed at low Péclet numbers, whereas wavy core-annular flows are obtained at moderate Reynolds numbers for both immiscible and miscible fluids. This work provides a unifying description of the influence of interfacial properties on viscous microflow phenomena.

Factors associated with the duration of hospitalisation among COVID-19 patients in Vietnam: A survival analysis.

BACKGROUND: The median duration of hospital stays due to COVID-19 has been reported in several studies on China as 10-13 days. Global studies have indicated that the length of hospitalisation depends on different factors, such as the time elapsed from exposure to symptom onset, and from symptom onset to hospital admission, as well as specificities of the country under study. The goal of this paper is to identify factors associated with the median duration of hospital stays of COVID-19 patients during the second COVID-19 wave that hit Vietnam from 5 March to 8 April 2020. METHOD: We used retrospective data on 133 hospitalised patients with COVID-19 recorded over at least two weeks during the study period. The Cox proportional-hazards regression model was applied to determine the potential risk factors associated with length of hospital stay. RESULTS: There were 65 (48.9%) females, 98 (73.7%) patients 48 years old or younger, 15 (11.3%) persons with comorbidities, 21 (16.0%) severely ill patients and 5 (3.8%) individuals with life-threatening conditions. Eighty-two (61.7%) patients were discharged after testing negative for the SARS-CoV-2 virus, 51 were still in the hospital at the end of the study period and none died. The median duration of stay in a hospital was 21 (IQR: 16-34) days. The multivariable Cox regression model showed that age, residence and sources of contamination were significantly associated with longer duration of hospitalisation. CONCLUSION: A close look at how long COVID-19 patients stayed in the hospital could provide an overview of their treatment process in Vietnam, and support the country's National Steering Committee on COVID-19 Prevention and Control in the efficient allocation of resources over the next stages of the COVID-19 prevention period.

Knowledge, Attitude, and Practice Regarding Breast Cancer Early Detection Among Women in a Mountainous Area in Northern Vietnam.

Breast cancer is the most common cancer in women all over the world, also in Vietnam. In recent years, the incidence of breast cancer has been increasing in Vietnam, and most cases are diagnosed at late stages, making treatment more difficult. More and better early detection could help more women to survive. The aim of this study was to identify the current knowledge, attitude and practice about early detection of breast cancer as well as potential predictors of breast cancer screening among women aged 20 to 49 year in a mountainous commune in Thanh Hoa Province, Vietnam, in a largely ethnic Muong population. Women aged 20 to 49 years were selected by systematic random sampling to participate in a cross sectional study in October 2017. They were interviewed with a closed questionnaire about their knowledge of breast cancer, its risk factors, and warning signs. A checklist for performance of breast self-examination was also applied. Three hundred six women agreed to participate in the study. More than half had a low level of knowledge, and were weak in attitude and practice about breast self-examination, clinical breast examination, breast ultrasound, and mamography. Among women who had practiced at least 1 screening method, 17.0% mentioned clinical breast examination, and only 13.8% reported practicing breast self-examination. Factors associated with practice included knowledge about breast cancer early detection (BCED), ethnicity, income, the BCED information approach, and the BCED screening programs approach. The finding of a very low proportion of women in the mountainous setting with good awareness and practice on early detection of breast cancer is important evidence to inform the BCED intervention program developers about where and how to target which information, especially to reach more ethnic minority women.

Study Design for the 2016 Baseline Survey of a Health System Strengthening Project in Quoc Oai District, Hanoi, Vietnam.

BACKGROUND: In order to provide essential scientific evidence on the population's health status and social health determinants as well as the current capacity of the health care system in Vietnam to health policy makers and managers, Vietnam Ministry of Health, Hanoi University of Public Health, Hanoi Medical University, and Ho Chi Minh University of Medicine and Pharmacy collaborated with Seoul National University (Korea) and conducted a health system survey in the Quoc Oai district (of Hanoi capital) that represented northern rural Vietnam. METHODS: The study design was a cross-sectional study. The survey covered different topics (more than 200 questions) and was administered in three separate questionnaires: 1) Basic information of all household members; 2) Household characteristics; and 3) Individual characteristics. Socio-demographic characteristics among the households and individuals were collected from 2,400 households sampled by multi-stage cluster sampling method: more than 200 questions. RESULTS: The household size of Quoc Oai was larger than the national average and there was no significant difference in gender composition. In addition, the proportions of pre-elderly, age 55-64, and elderly group (65 years old and over) were higher than the national population statistics. In this context, demographic transition has begun in Quoc Oai. CONCLUSION: This study design description provides the basic information about a baseline survey of a future prospective cohort (as a part of a collaborative project on strengthening the health system in Vietnam) to the prospective data user of this survey.

The roles of, activities of, and competencies for, community nursing services in rural Vietnam: Implications for policy decisions.

Community health workforce plays a vital role in providing primary health care services as per the needs of residents; however, few studies have examined how nurses work within commune health centers (CHCs). Using qualitative methods including interviews and focus group discussions with key stakeholders, this study explores the roles, activities, and competencies required of community nursing services in rural districts within Vietnam. Two primary roles were identified: CHC nursing and family nursing. For the latter, in addition to providing people with general health care and health communication, they were expected to also deliver psychological care. CHC nursing fulfilled more roles and required four specific competencies: clinical care, communication, management, and planning/coordination activities. Despite these various roles serving people within a community, few ongoing efforts at either the local or national level are aimed at supporting these nurses. The study highlights the need for policy decisions via either developing a new job position policy or adapting the existing policy by integrating new roles into the existing positions of CHC nurses in Vietnam.

Health Service Utilization Among People With Noncommunicable Diseases in Rural Vietnam.

CONTEXT: Chronic noncommunicable diseases (NCDs) have been shown to be major causes of morbidity and mortality in hospitals for the whole country. OBJECTIVE: This study aims to describe the situation of health service utilization among people with NCDs in a rural area and identify association between the situation of health service utilization among people with chronic diseases and their socioeconomic status. DESIGN: This was a cross-sectional study. SETTING: A rural district located in the North of Vietnam. PARTICIPANTS: People 15 years of age and older. Health service utilization was analyzed only among people who reported having NCD. MAIN OUTCOME MEASURES: Data were collected through a personal household interview conducted by 12 trained field workers. The dependent variable is health care service utilization among people with chronic NCDs. The explanatory variables include both household attributes such as household economic conditions, and so forth, and individual characteristics. RESULTS: Eighteen percent of the adults and 51% of the elderly respondents reported having at least 1 of the NCDs. The proportions of people with NCDs who used at least 1 outpatient service and used at least 1 inpatient health service during the last 12 months were 68.1% and 10.7%, respectively (the nonutilization rates of 31.9% and 89.3%, respectively). The statistically significant correlates of health care service utilization among people with NCDs were ethnicity (ethnic minority was significantly associated with a lower odds of health care service utilization) and health insurance (no health insurance was significantly associated with lower odds of health care service utilization). CONCLUSION: Given the evidence from this study, actions to improve access to health care services among people with NCDs are clearly needed. The capacity of primary health care system for the prevention and control of NCDs should be ranked a top priority.

La autoantigen mediates oxidant induced de novo Nrf2 protein translation.

Nrf2 gene encodes a transcription factor that regulates the expression of a cluster of antioxidant and detoxification genes. Recent works from our laboratory indicate that oxidative stress causes rapid de novo synthesis of Nrf2 protein. We have found that 5' Untranslated Region (5'UTR) of Nrf2 allows the mRNA to undergo an Internal Ribosomal Entry Site (IRES) mediated protein translation. Using liquid chromatography tandem MS, we have discovered that La/SSB protein bound to Nrf2 5'UTR in response to oxidative stress. In vitro RNA binding and in vivo ribonucleoprotein immunoprecipitation showed H(2)O(2) dose and time dependent increases of La/SSB binding to Nrf2 5'UTR. La/SSB protein translocated from the nuclei to cytoplasm and distributed in the perinuclear space in cells treated with H(2)O(2). Isolation of ribosomal fractions indicated that oxidants caused an association of La/SSB with ribosomes. Physical interaction of La/SSB with representative proteins from the small or large subunits of ribosomes was found to increase in cells responding to H(2)O(2) treatment. Knocking down La/SSB gene with siRNA prevented Nrf2 protein elevation or Nrf2 5'UTR activation by oxidants. In contrast, overexpression of La/SSB gene was able to enhance Nrf2 5'UTR activation and Nrf2 protein increase. Our data suggest that oxidants cause nuclear export of La/SSB protein and subsequent association of La/SSB with Nrf2 5'UTR and ribosomes. These events contribute to de novo Nrf2 protein translation because of oxidative stress.

The Message of the Glasgow Coma Scale: A Comprehensive Bibliometric Analysis and Systematic Review of Clinical Practice Guidelines Spanning the Past 50 years.

BACKGROUND: Despite the ubiquitous use of the Glasgow Coma Scale (GCS) worldwide, no study to date has objectively and quantitatively assessed its impact on the scientific literature and clinical practice. Therefore, we comprehensively analyzed scientific publications and clinical practice guidelines employing the GCS to gauge its clinical and academic impact, identify research hotspots, and inform future research on the topic. METHODS: A cross-sectional bibliometric analysis was performed on Scopus to obtain relevant publications incorporating the GCS from 1974 to 2022. In addition, a systematic review of existing clinical practice guidelines in PubMed, Scopus, Web of Science, and Trip Database was performed. Validated bibliometric parameters including article title, journal, publication year, authors, citation count, country, institution, keywords, impact factor, and references were assessed. When evaluating clinical practice guidelines, the sponsoring organization, country of origin, specialty, and publication year were assessed. RESULTS: A total of 37,633 articles originating from 3924 different scientific journals spanning 1974-2022 were included in the final analysis. The compound annual growth rate of publications referencing the GCS was 16.7%. Of 104 countries, the United States had the highest total number of publications employing the GCS (n = 8517). World Neurosurgery was the scientific periodical with the highest number of publications on the GCS (n = 798). The top trending author-supplied keyword was "traumatic brain injury" (n = 3408). The 97 included clinical practice guidelines most commonly employed the GCS in the fields of internal medicine (n = 22, 23%), critical care (n = 21, 22%), and neurotrauma (n = 19, 20%). CONCLUSIONS: At the turn of the 50th anniversary of the GCS, we provided a unique and detailed description of the "path to success" of the GCS both in terms of its scientific and clinical impact. These results have not only a historical but also an important didactic value. Ultimately our detailed analysis, which revealed some of the factors that led the GCS to become such a widespread and highly influential score, may assist future researchers in their development of new outcome measures and clinical scores, especially as such tools become increasingly relevant in an evidence-based data-driven age.

Computational tools supporting known miRNA identification.

The study of small RNAs is a field that is expanding quickly. Other functional short RNA molecules other than microRNAs, and gene expression regulators, have been found in animals and plants. MicroRNAs play a significant role in host-microbe interactions, and parasite microRNAs may affect the host's innate immunity. Furthermore, short RNAs are intriguing non-invasive biomarker possibilities because they can be found in physiological fluids. These trends suggest that for many researchers, quick and simple techniques for expression profiling and subsequent downstream analysis of miRNA-seq data are crucial. We selected sRNAtoolbox to make integrated sRNA research easier. Each tool can be used separately or to explore and analyze sRNAbench results in further depth. A special focus was placed on the tools' usability. We review available miRNA research tools to have an overview of the evaluation of the tools. Mainly we evaluate the tool sRNAtoolbox.

Microflow of nanoemulsion threads in surfactant solutions.

We experimentally investigate the microfluidic flow of oil-in-water nanoemulsions in aqueous sodium dodecyl sulfate (SDS) solutions having different concentrations and injection flow rates. A coaxial microfluidic device is employed to explore the behavior of nanoemulsion threads in these sheathing SDS solutions. Using two high-speed cameras, which simultaneously capture both top and side views, we reveal a variety of flow phenomena, ranging from simple core-annular flow to complex flows, such as gravitational, inertial, and buckling thread flows. By analyzing these complex flows, we develop a methodology that elucidates the relationship of core-annular and gravitational flows at low flow rates. Further, we examine the off-axis displacements and bending of core threads at large flow rates, and we study the buckling dynamics of nanoemulsion threads subjected to osmotic stresses caused by large SDS concentrations in the sheathing fluid.

Deep Generative Learning Models for Cloud Intrusion Detection Systems.

Intrusion detection (ID) on the cloud environment has received paramount interest over the last few years. Among the latest approaches, machine learning-based ID methods allow us to discover unknown attacks. However, due to the lack of malicious samples and the rapid evolution of diverse attacks, constructing a cloud ID system (IDS) that is robust to a wide range of unknown attacks remains challenging. In this article, we propose a novel solution to enable robust cloud IDSs using deep neural networks. Specifically, we develop two deep generative models to synthesize malicious samples on the cloud systems. The first model, conditional denoising adversarial autoencoder (CDAAE), is used to generate specific types of malicious samples. The second model (CDAEE-KNN) is a hybrid of CDAAE and the K -nearest neighbor algorithm to generate malicious borderline samples that further improve the accuracy of a cloud IDS. The synthesized samples are merged with the original samples to form the augmented datasets. Three machine learning algorithms are trained on the augmented datasets and their effectiveness is analyzed. The experiments conducted on four popular IDS datasets show that our proposed techniques significantly improve the accuracy of the cloud IDSs compared with the baseline technique and the state-of-the-art approaches. Moreover, our models also enhance the accuracy of machine learning algorithms in detecting some currently challenging distributed denial of service (DDoS) attacks, including low-rate DDoS attacks and application layer DDoS attacks.

Microplastics in sediments from urban and suburban rivers: Influence of sediment properties.

Although sediments are considered to be a major sink for microplastics (MP), there is still a relative lack of knowledge on the factors that influence the occurrence and abundance of MP in riverine sediments. The present study investigated the occurrence and distribution of MP in riverine sediments collected at twelve sites representative of different populated and urbanized rivers (To Lich, Nhue and Day Rivers) located in the Red River Delta (RRD, Vietnam, during dry and rainy seasons. MP concentrations ranged from 1600 items kg-1 dw to 94,300 items kg-1dw. Fiber shape dominated and MP were made of polypropylene (PP) and polyethylene (PE) predominantly. An absence of seasonal effect was observed for both fragments and fibers for each rivers. Decreasing MP concentrations trend was evidenced from the To Lich River, to the Nhue River and to the Day River, coupled with a decreasing fiber length and an increasing fragment area in the surface sediment from upstream to downstream. Content of organic matter was correlated to MP concentrations suggesting that, high levels of organic matter could be MP hotspots in urban rivers. Also, high population density as well as in highly residential areas are related to higher MP concentrations in sediments. Finally, a MP high ecological risk (RI: 866 to 4711) was calculated in the RDD.

Characterization of a high-affinity human antibody with a disulfide bridge in the third complementarity-determining region of the heavy chain.

Disulfide bridges are common in the antigen-binding site from sharks (new antigen receptor) and camels (single variable heavy-chain domain, VHH), in which they confer both structural diversity and domain stability. In human antibodies, cysteine residues in the third complementarity-determining region of the heavy chain (CDR-H3) are rare but naturally encoded in the IGHD germline genes. Here, by panning a phage display library designed based on human germline genes and synthetic CDR-H3 regions against a human cytokine, we identified an antibody (M3) containing two cysteine residues in the CDR-H3. It binds the cytokine with high affinity (0.4 nM), recognizes a unique epitope on the antigen, and has a distinct neutralization profile as compared with all other antibodies selected from the library. The two cysteine residues form a disulfide bridge as determined by mass spectrometric peptide mapping. Replacing the cysteines with alanines did not change the solubility and stability of the monoclonal antibody, but binding to the antigen was significantly impaired. Three-dimensional modeling and dynamic simulations were employed to explore how the disulfide bridge influences the conformation of CDR-H3 and binding to the antigen. On the basis of these results, we envision that designing human combinatorial antibody libraries to contain intra-CDR or inter-CDR disulfide bridges could lead to identification of human antibodies with unique binding profiles.

Commercial Potential of the Cyanobacterium Arthrospira maxima: Physiological and Biochemical Traits and the Purification of Phycocyanin.

Arthrospira maxima is a natural source of fine chemicals for multiple biotechnological applications. We determined the optimal environmental conditions for A. maxima by measuring its relative growth rate (RGR), pigment yield, and photosynthetic performance under different pH and temperature conditions. RGR was highest at pH 7-9 and 30 °C. Chlorophyll a, phycocyanin, maximal quantum yield (Fv/Fm), relative maximal electron transport rate (rETRmax), and effective quantum yield (ΦPSII) were highest at pH 7-8 and 25 °C. Interestingly, phycoerythrin and allophycocyanin content was highest at 15 °C, which may be the lowest optimum temperature reported for phycobiliprotein production in the Arthrospira species. A threestep purification of phycocyanin (PC) by ultrafiltration, ion-exchange chromatography, and gel filtration resulted in a 97.6% purity of PC.

Learning Latent Representation for IoT Anomaly Detection.

Internet of Things (IoT) has emerged as a cutting-edge technology that is changing human life. The rapid and widespread applications of IoT, however, make cyberspace more vulnerable, especially to IoT-based attacks in which IoT devices are used to launch attack on cyber-physical systems. Given a massive number of IoT devices (in order of billions), detecting and preventing these IoT-based attacks are critical. However, this task is very challenging due to the limited energy and computing capabilities of IoT devices and the continuous and fast evolution of attackers. Among IoT-based attacks, unknown ones are far more devastating as these attacks could surpass most of the current security systems and it takes time to detect them and "cure" the systems. To effectively detect new/unknown attacks, in this article, we propose a novel representation learning method to better predictively "describe" unknown attacks, facilitating supervised learning-based anomaly detection methods. Specifically, we develop three regularized versions of autoencoders (AEs) to learn a latent representation from the input data. The bottleneck layers of these regularized AEs trained in a supervised manner using normal data and known IoT attacks will then be used as the new input features for classification algorithms. We carry out extensive experiments on nine recent IoT datasets to evaluate the performance of the proposed models. The experimental results demonstrate that the new latent representation can significantly enhance the performance of supervised learning methods in detecting unknown IoT attacks. We also conduct experiments to investigate the characteristics of the proposed models and the influence of hyperparameters on their performance. The running time of these models is about 1.3 ms that is pragmatic for most applications.

Progression of liver fibrosis and associated factors among chronic hepatitis B patients at a general hospital in Northern Vietnam.

Evaluation of liver fibrosis is necessary to make the therapeutic decision and assess the prognosis of CHB patients. The current study aimed to describe the progression and identify some influencing factors in patients with chronic hepatitis B at a General Hospital in Northern Vietnam. The longitudinal study included 55 eligible subjects diagnosed Hepatitis-B-virus. Dependent variable was the aspartate aminotransferase/platelet ratio index and we collected some demographic variables and disease related and behaviour variables. Bayesian Model Averaging was used to select variables into model. Mixed-effect linear models were used to evaluate the change of the aspartate aminotransferase/platelet ratio index over time and identify related factors. the aspartate aminotransferase/platelet ratio index differences between examinations, age of participants, working status were statistically significant. This pattern indicated that the average the aspartate aminotransferase/platelet ratio index of the population decreased by 0.005 (95% CI=-0.009; -0.001) after each patient's visit, and increased by 0.013 if the patient's age increased by 1 year (95% CI=0.005; 0.0219). For non-working patients, the aspartate aminotransferase/platelet ratio index was lower, coefficient was -0.054 (95% CI=-0.108; 0.001). Other variables such as gender, education level, time for disease detection, drinking tea, alcohol consumption, forgetting to take medicine and the aspartate aminotransferase/platelet ratio index were not significantly different. The study showed that the majority of study subjects had average the aspartate aminotransferase/platelet ratio index, and were relatively well controlled and treated during the study. Age and working status are factors that influence the the aspartate aminotransferase/platelet ratio index.

p21(WAF1/Cip1/Sdi1) knockout mice respond to doxorubicin with reduced cardiotoxicity.

Doxorubicin (Dox) is an antineoplastic agent that can cause cardiomyopathy in humans and experimental animals. As an inducer of reactive oxygen species and a DNA damaging agent, Dox causes elevated expression of p21(WAF1/Cip1/Sdi1) (p21) gene. Elevated levels of p21 mRNA and p21 protein have been detected in the myocardium of mice following Dox treatment. With chronic treatment of Dox, wild type (WT) animals develop cardiomyopathy evidenced by elongated nuclei, mitochondrial swelling, myofilamental disarray, reduced cardiac output, reduced ejection fraction, reduced left ventricular contractility, and elevated expression of ANF gene. In contrast, p21 knockout (p21KO) mice did not show significant changes in the same parameters in response to Dox treatment. In an effort to understand the mechanism of the resistance against Dox induced cardiomyopathy, we measured levels of antioxidant enzymes and found that p21KO mice did not contain elevated basal or inducible levels of glutathione peroxidase and catalase. Measurements of 6 circulating cytokines indicated elevation of IL-6, IL-12, IFNγ and TNFα in Dox treated WT mice but not p21KO mice. Dox induced elevation of IL-6 mRNA was detected in the myocardium of WT mice but not p21KO mice. While the mechanism of the resistance against Dox induced cardiomyopathy remains unclear, lack of inflammatory response may contribute to the observed cardiac protection in p21KO mice.

Unmet Need for Emergency Medical Services in Hanoi, Vietnam.

Low- and middle-income countries urgently need to improve emergency medical services (EMSs) as a component of their healthcare systems. Here, we detailed EMS resources and their provision in Hanoi, Vietnam, and discussed necessary policies to upgrade EMSs. Between 2013 and 2018, EMS resources, measured as provider-to-population and ambulance-to-population ratios, decreased, whereas service provision, measured as the number of patients transported by ambulance per population, increased. EMS resources and their provision in Hanoi are far below the standards of high-income countries or figures in neighboring Asian countries. Therefore, it is imperative to upgrade health policies for the appropriate allocation of healthcare resources to EMSs and hospital services.

Far Upstream Binding Protein 1 (FUBP1) participates in translational regulation of Nrf2 protein under oxidative stress.

Oxidative stress is ubiquitously involved in disease etiology or progression. While the damaging effects have been well characterized, how cells deal with oxidative stress for prevention or removal of damage remains to be fully elucidated. Works from our laboratory have revealed de novo Nrf2 protein translation when cells are encountering low to mild levels of oxidative stress. Nrf2 encodes a transcription factor controlling a myriad of genes important for antioxidation, detoxification, wound repair and tissue remodeling. Here we report a role of FUBP1 in regulating de novo Nrf2 protein translation. An increase of FUBP1 binding to Nrf2 5'UTR due to H2O2 treatment has been found by LC-MS/MS, Far Western blot and ribonucleoprotein immunoprecipitation assays. Blocking FUBP1 expression using siRNA abolished H2O2 from inducing Nrf2 protein elevation or Nrf2 5'UTR activity. While no nuclear to cytoplasmic translocation was detected, cytosolic redistribution to the ribosomal fractions was observed due to oxidant treatment. The presence of FUBP1 in 40/43S ribosomal fractions confirm its involvement in translation initiation of Nrf2 protein. When tested by co-immunoprecipitation with eIF4E, eIF2a, eIF3η and eIF1, only eIF3η was found to gain physical interaction with FUBP1 due to H2O2 treatment. Our data support a role of FUBP1 for promoting the attachment of 40S ribosomal subunit to Nrf2 mRNA and formation of 43S pre-initiation complex for translation initiation of Nrf2 protein under oxidative stress.

An automated alarm system for food safety by using electronic invoices.

BACKGROUND: Invoices had been used in food product traceability, however, none have addressed the automated alarm system for food safety by utilizing electronic invoice big data. In this paper, we present an alarm system for edible oil manufacture that can prevent a food safety crisis rather than trace problematic sources post-crisis. MATERIALS AND METHODS: Using nearly 100 million labeled e-invoices from the 2013‒2014 of 595 edible oil manufacturers provided by Ministry of Finance, we applied text-mining, statistical and machine learning techniques to "train" the system for two functions: (1) to sieve edible oil-related e-invoices of manufacturers who may also produce other merchandise and (2) to identify suspicious edible oil manufacture based on irrational transactions from the e-invoices sieved. RESULTS: The system was able to (1) accurately sieve the correct invoices with sensitivity >95% and specificity >98% via text classification and (2) identify problematic manufacturers with 100% accuracy via Random Forest machine learning method, as well as with sensitivity >70% and specificity >99% through simple decision-tree method. CONCLUSION: E-invoice has bright future on the application of food safety. It can not only be used for product traceability, but also prevention of adverse events by flag suspicious manufacturers. Compulsory usage of e-invoice for food producing can increase the accuracy of this alarm system.