Usefulness of Vaccine Adverse Event Reporting System for Machine-Learning Based Vaccine Research: A Case Study for COVID-19 Vaccines.

Usefulness of Vaccine-Adverse Event-Reporting System (VAERS) data and protocols required for statistical analyses were pinpointed with a set of recommendations for the application of machine learning modeling or exploratory analyses on VAERS data with a case study of COVID-19 vaccines (Pfizer-BioNTech, Moderna, Janssen). A total of 262,454 duplicate reports (29%) from 905,976 reports were identified, which were merged into a total of 643,522 distinct reports. A customized online survey was also conducted providing 211 reports. A total of 20 highest reported adverse events were first identified. Differences in results after applying various machine learning algorithms (association rule mining, self-organizing maps, hierarchical clustering, bipartite graphs) on VAERS data were noticed. Moderna reports showed injection-site-related AEs of higher frequencies by 15.2%, consistent with the online survey (12% higher reporting rate for pain in the muscle for Moderna compared to Pfizer-BioNTech). AEs {headache, pyrexia, fatigue, chills, pain, dizziness} constituted >50% of the total reports. Chest pain in male children reports was 295% higher than in female children reports. Penicillin and sulfa were of the highest frequencies (22%, and 19%, respectively). Analysis of uncleaned VAERS data demonstrated major differences from the above (7% variations). Spelling/grammatical mistakes in allergies were discovered (e.g., ~14% reports with incorrect spellings for penicillin).

Cross Lingual Sentiment Analysis: A Clustering-Based Bee Colony Instance Selection and Target-Based Feature Weighting Approach.

The lack of sentiment resources in poor resource languages poses challenges for the sentiment analysis in which machine learning is involved. Cross-lingual and semi-supervised learning approaches have been deployed to represent the most common ways that can overcome this issue. However, performance of the existing methods degrades due to the poor quality of translated resources, data sparseness and more specifically, language divergence. An integrated learning model that uses a semi-supervised and an ensembled model while utilizing the available sentiment resources to tackle language divergence related issues is proposed. Additionally, to reduce the impact of translation errors and handle instance selection problem, we propose a clustering-based bee-colony-sample selection method for the optimal selection of most distinguishing features representing the target data. To evaluate the proposed model, various experiments are conducted employing an English-Arabic cross-lingual data set. Simulations results demonstrate that the proposed model outperforms the baseline approaches in terms of classification performances. Furthermore, the statistical outcomes indicate the advantages of the proposed training data sampling and target-based feature selection to reduce the negative effect of translation errors. These results highlight the fact that the proposed approach achieves a performance that is close to in-language supervised models.

Pupil Localisation and Eye Centre Estimation Using Machine Learning and Computer Vision.

Various methods have been used to estimate the pupil location within an image or a real-time video frame in many fields. However, these methods lack the performance specifically in low-resolution images and varying background conditions. We propose a coarse-to-fine pupil localisation method using a composite of machine learning and image processing algorithms. First, a pre-trained model is employed for the facial landmark identification to extract the desired eye frames within the input image. Then, we use multi-stage convolution to find the optimal horizontal and vertical coordinates of the pupil within the identified eye frames. For this purpose, we define an adaptive kernel to deal with the varying resolution and size of input images. Furthermore, a dynamic threshold is calculated recursively for reliable identification of the best-matched candidate. We evaluated our method using various statistical and standard metrics along with a standardised distance metric that we introduce for the first time in this study. The proposed method outperforms previous works in terms of accuracy and reliability when benchmarked on multiple standard datasets. The work has diverse artificial intelligence and industrial applications including human computer interfaces, emotion recognition, psychological profiling, healthcare, and automated deception detection.

Health for all by all-pursuing multi-sectoral action on health for SDGs in the WHO Eastern Mediterranean Region.

The WHO Eastern Mediterranean Region is endowed with deep intellectual tradition, interesting cultural diversity, and a strong societal fabric; components of a vibrant platform for promoting health and wellbeing. Health has a central place in the Sustainable Development Goals (SDGs) for at least three reasons: Firstly, health is shaped by factors outside of the health sector. Secondly, health can be singled out among several SDGs as it provides a clear lens for examining the progress of the entire development process. Thirdly, in addition to being an outcome, health is also a contributor to achieving sustainable development. Realizing this central role of health in SDGs and the significance of collaboration among diverse sectors, the WHO is taking action. In its most recent General Program of Work 2019-2023 (GPW 13), the WHO has set a target of promoting the health of one billion more people by addressing social and other determinants of health through multi-sectoral collaboration. The WHO Regional Office for the Eastern Mediterranean Region, through Vision 2023, aims at addressing these determinants by adopting an equity-driven, leaving no one behind approach. Advocating for Health in All Policies, multi-sectoral action, community engagement, and strategic partnerships are the cornerstone for this approach. The focus areas include addressing the social and economic determinants of health across the life course, especially maternal and child health, communicable diseases, non-communicable diseases, and injuries. The aspirations are noteworthy - however, recent work in progress in countries has also highlighted some areas for improvement. Joint work among different ministries and departments at country level is essential to achieve the agenda of sustainable development. For collaboration, not only the ministries and departments need to be engaged, but the partnerships with other stakeholders such as civil society and private sector are a necessity and not a choice to effectively pursue achievement of SDGs.

Influenza in the Eastern Mediterranean Region: identifying the unknowns for detection and control of epidemic and pandemic threats (Editorial).

Acute respiratory illnesses and influenza-like illnesses (ILI) are a significant cause of morbidity and mortality worldwide. Data from developed countries reveal that seasonal influenza can affect up to 15% of the population presenting with upper respiratory tract infections and may result in up to 500 000 deaths worldwide annually. Despite their public health importance, little was known about the aetiology of these illnesses in the countries of the WHO Eastern Mediterranean Region (EMR).

Research agenda on persistent and unpredictable threat of influenza and emerging respiratory infections: a public health necessity in the Eastern Mediterranean Region (Guest Editorial).

Infectious diseases continue to represent a significant threat to global health security, particularly in the context of increasing globalization, interconnectedness and interdependence. Chief among such threats are influenza viruses and other respiratory pathogens, such as Middle East Respiratory Syndrome coronavirus (MERS-CoV), because of their risk of high transmissibility and acuity of illness. Annual epidemics of seasonal influenza cause an estimated 3-5 million cases of severe illness and more than 500 000 deaths, with the prospect of pandemic influenza viruses causing far greater impact. In addition, the appearance of severe acute respiratory syndrome (SARS) in 2003, widespread and continued outbreaks of avian influenza A (H5N1) since 2004, the H1N1 pandemic in 2009 and emergence of MERS CoV in 2012 reflect the seriousness of public health challenges posed by influenza and emerging respiratory infections.

RISK CHARACTERIZATION OF MATERNAL AND NEONATAL TETANUS IN VIEW OF TETANUS VACCINATION CAMPAIGNS IN PAKISTAN.

BACKGROUND: Pakistan is one of the remaining 24 countries which have not yet achieved Maternal and Neonatal Tetanus Elimination (MNTE), The country adopted high-risk approach for 56 out of 119 districts with country-wide Tetanus Toxoid (TT) provision in Routine Immunization (RI) during early 2000-2003. The TT's mass campaigns could only cover 13% of high risk districts for 2009- 2011, and mostly for the Punjab province. To achieve MNT elimination, the country needs risk mapping for cost-effective intervention. METHODS: We used both the quantitative and qualitative methods to conduct risk characterization. All the three available data sets (Reported EPI coverage data, PDHS 2012-13, and PSLM 2010-11) were assessed. A mix of core and surrogate indicators-for risk categorization was used through ranking and scoring the aggregated data and considering the past tetanus campaigns' coverage. Tetanus Toxoid (TT2+)-coverage of pregnant women and delivery in health facility, both received more weightage in scoring. We based the higher and lower cuts off points for each indicator on data ranges. The districts with higher scores, i.e., 10.5 and above were ranked good followed by medium (5.5-10.4) and low performing (less than 5.5). Consultations with the national and provincial field officers were utilized to understand the local context. RESULTS: In Pakistan, there are 139 districts out of which, 60 are the high risk districts for tetanus. Highest percentage is for Baluchistan (83%) followed by Sindh (52%), and Khyber Pakhtunkhwa (40%). Most of the Punjab is at medium risk (55%), followed by KP (52%), and Sindh (39%). CONCLUSION: Pakistan is at medium to high risk of MNT with a great variation at the sub-national level. Campaigns aiming to these districts may bring the country closer to MNT elimination target.

An explainable machine learning approach for Alzheimer's disease classification.

The early diagnosis of Alzheimer's disease (AD) presents a significant challenge due to the subtle biomarker changes often overlooked. Machine learning (ML) models offer a promising tool for identifying individuals at risk of AD. However, current research tends to prioritize ML accuracy while neglecting the crucial aspect of model explainability. The diverse nature of AD data and the limited dataset size introduce additional challenges, primarily related to high dimensionality. In this study, we leveraged a dataset obtained from the National Alzheimer's Coordinating Center, comprising 169,408 records and 1024 features. After applying various steps to reduce the feature space. Notably, support vector machine (SVM) models trained on the selected features exhibited high performance when tested on an external dataset. SVM achieved a high F1 score of 98.9% for binary classification (distinguishing between NC and AD) and 90.7% for multiclass classification. Furthermore, SVM was able to predict AD progression over a 4-year period, with F1 scores reached 88% for binary task and 72.8% for multiclass task. To enhance model explainability, we employed two rule-extraction approaches: class rule mining and stable and interpretable rule set for classification model. These approaches generated human-understandable rules to assist domain experts in comprehending the key factors involved in AD development. We further validated these rules using SHAP and LIME models, underscoring the significance of factors such as MEMORY, JUDGMENT, COMMUN, and ORIENT in determining AD risk. Our experimental outcomes also shed light on the crucial role of the Clinical Dementia Rating tool in predicting AD.

Anticancer drug delivery: Investigating the impacts of viscosity on lipid-based formulations for pulmonary targeting.

Pulmonary drug delivery via aerosolization is a non-intrusive method for achieving localized and systemic effects. The aim of this study was to establish the impact of viscosity as a novel aspect (i.e., low, medium and high) using various lipid-based formulations (including liposomes (F1-F3), transfersomes (F4-F6), micelles (F7-F9) and nanostructured lipid carriers (NLCs; F10-F12)) as well as to investigate their impact on in-vitro nebulization performance using Trans-resveratrol (TRES) as a model anticancer drug. Based on the physicochemical properties, micelles (F7-F9) elicited the smallest particle size (12-174 nm); additionally, all formulations tested exhibited high entrapment efficiency (>89 %). Through measurement using capillary viscometers, NLC formulations exhibited the highest viscosity (3.35-10.04 m2/sec). Upon using a rotational rheometer, formulations exhibited shear-thinning (non-Newtonian) behaviour. Air jet and vibrating mesh nebulizers were subsequently employed to assess nebulization performance using an in-vitro model. Higher viscosity formulations elicited a prolonged nebulization time. The vibrating mesh nebulizer exhibited significantly higher emitted dose (ED), fine particle fraction (FPF) and fine particle dose (FPD) (up to 97 %, 90 % and 64 µg). Moreover, the in-vitro release of TRES was higher at pH 5, demonstrating an alignment of the release profile with the Korsmeyer-Peppas model. Thus, formulations with higher viscosity paired with a vibrating mesh nebulizer were an ideal combination for delivering and targeting peripheral lungs.

Video dataset of sheep activity for animal behavioral analysis via deep learning.

A primary dataset capturing five distinct types of sheep activities in realistic settings was constructed at various resolutions and viewing angles, targeting the expansion of the domain knowledge for non-contact virtual fencing approaches. The present dataset can be used to develop non-invasive approaches for sheep activity detection, which can be proven useful for farming activities including, but not limited to, sheep counting, virtual fencing, behavior detection for health status, and effective sheep breeding. Sheep activity classes include grazing, running, sitting, standing, and walking. The activities of individuals, as well as herds of sheep, were recorded at different resolutions and angles to provide a dataset of diverse characteristics, as summarized in Table 1. Overall, a total of 149,327 frames from 417 videos (the equivalent of 59 minutes of footage) are presented with a balanced set for each activity class, which can be utilized for robust non-invasive detection models based on computer vision techniques. Despite a decent existence of noise within the original data (e.g., segments with no sheep present, multiple sheep in single frames, multiple activities by one or more sheep in single as well as multiple frames, segments with sheep alongside other non-sheep objects), we provide original videos and the original videos' frames (with videos and frames containing humans omitted for privacy reasons). The present dataset includes diverse sheep activity characteristics and can be useful for robust detection and recognition models, as well as advanced activity detection models as a function of time for the applications.

A review of policies and coverage of seasonal influenza vaccination programs in the WHO Eastern Mediterranean Region.

Background: Although there has been an effective seasonal influenza vaccine available for more than 60 years, influenza continues to circulate and cause illness. The Eastern Mediterranean Region (EMR) is very diverse in health systems capacities, capabilities, and efficiencies, which affect the performance of services, especially vaccination, including seasonal influenza vaccination. Aims: The aim of this study is to provide a comprehensive overview on country-specific influenza vaccination policies, vaccine delivery, and coverage in EMR. Materials and Methods: We have analyzed data from a regional seasonal influenza survey conducted in 2022, Joint Reporting Form (JRF), and verified their validity by the focal points. We also compared our results with those of the regional seasonal influenza survey conducted in 2016. Results: Fourteen countries (64%) had reported having a national seasonal influenza vaccine policy. About (44%) countries recommended influenza vaccine for all SAGE recommended target groups. Up to 69% of countries reported that COVID-19 had an impact on influenza vaccine supply in the country, with most of them (82%) reporting increases in procurement due to COVID-19. Discussion: The situation of seasonal influenza vaccination in EMR is varied, with some countries having well established programs while others having no policy or program; these variances may be due to resources inequity, political, and socioeconomic dissimilarities. Few countries have reported wide vaccination coverage over time with no clear trend of improvement. Conclusion: We suggest supporting countries to develop a roadmap for influenza vaccine uptake and utilization, assessment of barriers, and burden of influenza, including measuring the economic burden to enhance vaccine acceptance.

Rapid assessment of case recruitment tools to inform integrated surveillance of influenza and other respiratory viruses in Eastern Mediterranean countries.

Influenza-like illness (ILI) and severe acute respiratory infection (SARI) case recruitment tools from 10 countries were reviewed. The contents of the existing tools were compared against World Health Organization's current guidelines, and we also assessed the content validity (accuracy, completeness and consistency). Five of the ILI tools and two of the SARI tools were rated as having high accuracy against WHO case definitions. ILI completeness ranged from 25% to 86% and SARI from 52% to 96%. Average internal consistency scores were 86% for ILI and 94% for SARI. Limitations in the content validity of influenza case recruitment tools may compromise recruitment of eligible cases and result in varying detection rates across countries.

Wide and deep learning based approaches for classification of Alzheimer's disease using genome-wide association studies.

The increasing incidence of Alzheimer's disease (AD) has been leading towards a significant growth in socioeconomic challenges. A reliable prediction of AD might be useful to mitigate or at-least slow down its progression for which, identification of the factors affecting the AD and its accurate diagnoses, are vital. In this study, we use Genome-Wide Association Studies (GWAS) dataset which comprises significant genetic markers of complex diseases. The original dataset contains large number of attributes (620901) for which we propose a hybrid feature selection approach based on association test, principal component analysis, and the Boruta algorithm, to identify the most promising predictors of AD. The selected features are then forwarded to a wide and deep neural network models to classify the AD cases and healthy controls. The experimental outcomes indicate that our approach outperformed the existing methods when evaluated on standard dataset, producing an accuracy and f1-score of 99%. The outcomes from this study are impactful particularly, the identified features comprising AD-associated genes and a reliable classification model that might be useful for other chronic diseases.

Performance enhancement of high order Hahn polynomials using multithreading.

Orthogonal polynomials and their moments have significant role in image processing and computer vision field. One of the polynomials is discrete Hahn polynomials (DHaPs), which are used for compression, and feature extraction. However, when the moment order becomes high, they suffer from numerical instability. This paper proposes a fast approach for computing the high orders DHaPs. This work takes advantage of the multithread for the calculation of Hahn polynomials coefficients. To take advantage of the available processing capabilities, independent calculations are divided among threads. The research provides a distribution method to achieve a more balanced processing burden among the threads. The proposed methods are tested for various values of DHaPs parameters, sizes, and different values of threads. In comparison to the unthreaded situation, the results demonstrate an improvement in the processing time which increases as the polynomial size increases, reaching its maximum of 5.8 in the case of polynomial size and order of 8000 × 8000 (matrix size). Furthermore, the trend of continuously raising the number of threads to enhance performance is inconsistent and becomes invalid at some point when the performance improvement falls below the maximum. The number of threads that achieve the highest improvement differs according to the size, being in the range of 8 to 16 threads in 1000 × 1000 matrix size, whereas at 8000 × 8000 case it ranges from 32 to 160 threads.

A diverse and multi-modal gait dataset of indoor and outdoor walks acquired using multiple cameras and sensors.

Gait datasets are often limited by a lack of diversity in terms of the participants, appearance, viewing angle, environments, annotations, and availability. We present a primary gait dataset comprising 1,560 annotated casual walks from 64 participants, in both indoor and outdoor real-world environments. We used two digital cameras and a wearable digital goniometer to capture visual as well as motion signal gait-data respectively. Traditional methods of gait identification are often affected by the viewing angle and appearance of the participant therefore, this dataset mainly considers the diversity in various aspects (e.g., participants' attributes, background variations, and view angles). The dataset is captured from 8 viewing angles in 45° increments along-with alternative appearances for each participant, for example, via a change of clothing. The dataset provides 3,120 videos, containing approximately 748,800 image frames with detailed annotations including approximately 56,160,000 bodily keypoint annotations, identifying 75 keypoints per video frame, and approximately 1,026,480 motion data points captured from a digital goniometer for three limb segments (thigh, upper arm, and head).