Development of an Early Alert System for an Additional Wave of COVID-19 Cases Using a Recurrent Neural Network with Long Short-Term Memory

Abstract

The impact of the still ongoing “Coronavirus Disease 2019” (COVID-19) pandemic has been and is still vast, affecting not only global human health and stretching healthcare facilities, but also profoundly disrupting societal and economic systems worldwide. The nature of the way the virus spreads causes cases to come in further recurring waves. This is due a complex array of biological, societal and environmental factors, including the novel nature of the emerging pathogen. Other parameters explaining the epidemic trend consisting of recurring waves are logistic–organizational challenges in the implementation of the vaccine roll-out, scarcity of doses and human resources, seasonality, meteorological drivers, and community heterogeneity, as well as cycles of strengthening and easing/lifting of the mitigation interventions. Therefore, it is crucial to be able to have an early alert system to identify when another wave of cases is about to occur. The availability of a variety of newly developed indicators allows for the exploration of multi-feature prediction models for case data. Ten indicators were selected as features for our prediction model. The model chosen is a Recurrent Neural Network with Long Short-Term Memory. This paper documents the development of an early alert/detection system that functions by predicting future daily confirmed cases based on a series of features that include mobility and stringency indices, and epidemiological parameters. The model is trained on the intermittent period in between the first and the second wave, in all of the South African provinces.

A Bidirectional Long Short-Term Memory Model Algorithm for Predicting COVID-19 in Gulf Countries

“…Numerous studies have employed artificial intelligence models for estimating and predicting COVID-19. Stevenson et al [81] applied three models (LSTM, naïve, and seasonal naïve forecast), to predict COVID-19 and they used RMSE metrics to evaluate the obtained results for 7-day and 14-day time intervals to predict future daily cases in South Africa. Satu et al [82] proposed polynomial multi-layer perceptron (Poly-MLP), support vector regression (SVR), and Prophet models to predict confirmed and death cases using two evaluation metrics, RMSE and R2, to examine the proposed models.…”

Redefining pandemic preparedness: Multidisciplinary insights from the CERP modelling workshop in infectious diseases, workshop report

“…Jude Kong shared the experience of the “Africa-Canada Artificial Intelligence and Data Innovation Consortium” during Covid-19 and how this can be leveraged and capitalized for other respiratory diseases ( https://acadic.org/covid-19-dashboards/ ) ( Kong et al., 2023 ; Stevenson et al., 2021 ). The consortium, established in 2020 with a focus on fostering knowledge sharing, operates across ten African countries.…”

“…Their approach revolved around sourcing data from community-led organizations and partnering with telecommunications giants, Orange and MTN, to collect vital community-level information. Once data were amassed, they were processed into long- and short-term memory datasets, serving as the foundation for training a recurrent neural network model capable of predicting disease trends up to 14 days in advance ( Stevenson et al., 2021 ). This predictive model was then cross-referenced with historical outbreak data to establish a value threshold.…”

Leveraging Responsible, Explainable, and Local Artificial Intelligence Solutions for Clinical Public Health in the Global South

“…ACADIC has been devising and deploying AI-and BDA-based techniques to better understand the impacts of clinical public and global health interventions implemented during the COVID-19 pandemic in the Global South [85][86][87][88][89][90][91][92]. More specifically, ACADIC has been assisting policy-and decision-makers with the fight against COVID-19, including (i) monitoring and forecasting the growth and spread of COVID-19 at the local, state, and national levels [85,92], (ii) evaluating efforts to mitigate and control the spread [92], (iii) identifying trends in the disease infections, hospitalizations, and deaths [92], (iv) guiding purchase and allocation of health care resources [85], (v) guiding the collection of data (ensuring that data were disaggregated by race, gender, sexuality, class, geographic location, and Indigeneity) [92,93], (vi) guiding the implementation of vaccine roll-out and the development of effective, data-driven, evidence-informed immunization strategies, taking into account that available supply of vaccines in Africa was limited [87,90,91]; (vii) pro-viding situational intelligence: on populations at risk, stage of the outbreak, the projected burden of illness, school/business/work closure and re-opening, etc.…”

“…More specifically, ACADIC has been assisting policy-and decision-makers with the fight against COVID-19, including (i) monitoring and forecasting the growth and spread of COVID-19 at the local, state, and national levels [85,92], (ii) evaluating efforts to mitigate and control the spread [92], (iii) identifying trends in the disease infections, hospitalizations, and deaths [92], (iv) guiding purchase and allocation of health care resources [85], (v) guiding the collection of data (ensuring that data were disaggregated by race, gender, sexuality, class, geographic location, and Indigeneity) [92,93], (vi) guiding the implementation of vaccine roll-out and the development of effective, data-driven, evidence-informed immunization strategies, taking into account that available supply of vaccines in Africa was limited [87,90,91]; (vii) pro-viding situational intelligence: on populations at risk, stage of the outbreak, the projected burden of illness, school/business/work closure and re-opening, etc. [88], (viii) nowcasting labor market flow [89], (ix) supporting race, gender, sexuality, class, geographic location, and Indigeneity, inclusive COVID-19 actions [92], (x) developing methodologies and technologies to describe contact mixing and transmission networks to quantify impacts of contact shifting and individual mobility [92], (xi) supporting transparent and responsible AI, data, and digital rights governance around COVID-19 and pandemic responses [92], (xii) strengthening data systems and information sharing about COVID-19, (xiii) building trust and combatting mis-and dis-information around COVID-19 [91], (xiv) optimizing public health system responses for patient diagnosis, care, and management [92], (xv) establishing sustainable collaborations among model developers, policymakers, community leaders, etc. [92], (xvi) preparing the next generation of leaders in infectious disease AIand BDA-based modeling approaches in these countries [92], (xvii) working closely with public health agencies and other stakeholders to build trust and knowledge of AI-based models among key decision-makers [92], (xviii) developing stand-alone and predictive clinical public health decision support tools [92], and, (xix) creating a collaborative network that can respond rapidly to support decision-makers in each country to address infectious diseases or other disasters and emergency situations in general [92].…”