Leveraging Diagnostic data and AI for early outbreak detection
A disease outbreak and response system driven by AI predictive modeling and analysis will be developed. The system will be fed by real-time molecular diagnostic data, geospatial, seasonal and population level data to provide early warning for outbreaks (location/magnitude), forecasting of potential outbreak trajectories and modeling resources needed for containment.
Chris Macek, CEO, SystemOne, LLC
- Identify (Determine & limit the disease risk pool & spill over risk), such as: Genomic data to predict emerging risk, Early warning through ecological, behavioural & other data, Intervention/Incentives to reduce risk for emergency & spill over
Infectious disease outbreaks are increasing with a resurgence of epidemics such as cholera, Ebola and dengue, particularly in Africa. With an already existing burden of HIV and tuberculosis in low-income countries, now worsened by COVID-19, there is a need to leverage molecular diagnostics and genetic metadata to identify missing cases, track outbreaks and strain diversity of infectious agents.
Disease surveillance systems, data sharing and management are the weakest links in the health system of low-income countries, hindering governments’ ability to timely predict, prevent and respond to epidemics or tailor interventions. Although surveillance is crucial for understanding epidemiological trends for preparedness and response efforts, current systems are woefully unprepared to gather this information.
Three million people with TB are considered missing each year because of under-diagnosis and underreporting (WHO). During the 2014-2016 Ebola outbreak, only 60% patients were laboratory-confirmed and surveillance systems relied almost exclusively on error-prone manual/passive reporting. So too, the current COVID-19 pandemic has been challenged by data problems including poor data collection, low data quality, poor data usage and untimely data sharing.
This proposal aims to leverage the advances in IoT (Internet of Things), molecular diagnostics and AI breakthroughs to provide a real-time outbreak identification system.
It is anticipated that a centralised portal for disease results will allow Ministries of Health (MoH) in LMICs to rapidly detect transmission hotspots and prediction of new outbreaks through integration and deployment with local Emergency Operation Centers (EOC) and according to local response requirements. SystemOne will leverage existing relations established with the various MOH in order to deploy the solution to local contexts. In addition, it will enable National programs to leverage existing technology to rapidly increase their ability to respond appropriately, and ease coordination and planning for the available resources required to contain future outbreaks.
The proposed surveillance system will also directly address both the improvement of community-based care delivery and strengthen primary healthcare systems. Through implementation of Aspect Reporter, our mobile application, infectious disease diagnostic results can be automatically delivered digitally directly to healthcare providers on the frontline, in real-time, allowing prompt detection of patients requiring early treatment/quarantine, thus helping to limit the spread of the disease in the community. Positive results and patient contact information can also be automatically sent to contact tracing teams via Aspect to speed up community response, as well as to the patient, to ensure rapid isolation and behavioral change.
- Growth: An initiative, venture, or organisation with an established product, service, or business/policy model rolled out in one or, ideally, several contexts or communities, which is poised for further growth
- Artificial Intelligence / Machine Learning
- Big Data
- GIS and Geospatial Technology
- Internet of Things
- Software and Mobile Applications
The developed solution would be an additional product module which could be immediately leveraged by existing customers and included as part of the system offering for all new customers. In this context, Aspect and its associated reporter software through the provision of real time test results and linkages to patient intervention and outcomes, will limit the transmission of the target pathogen, reduces morbidity and mortality including worker absenteeism through rapid access to interventions thereby contributing to the overall improvement in the functioning of the health system.
The environmental and epidemiological twin model will be made available for scientific research and collaboration. 3rd party health partners can upload reference material for cause-effect analysis and augment patient level data for profiling and segmentation.
The Aspect system currently provides tangible impact in implemented countries to reduce result delivery turnaround times, improve operational logistics, increase initiation in care and overall cost-savings for programs of the national MoH (https://systemone.id/evidence-of-impact). SystemOne envision that the output from this project would further extend these benefits to national MoH’s by shortening the time from positive case detection to outbreak identification, and ultimately provide insights and recommendations to inform resources required to contain specific outbreaks of infectious diseases.
Using contextual real-world data, the platform will highlight vulnerable population groups and high-burden disease areas. Knowing the distribution and accessibility of health services within a given region, the model will provide insights as to the most effective use of resources and type of intervention. In addition, through focusing on major diseases of poverty and facilitating surveillance for potential outbreaks, vulnerable populations most at risk are protected, and this will inform the provision of targeted, early interventions. Furthermore, the ability to track outcomes and treatments provided through the Aspect Reporter and by linking this data to AI/ML-driven platforms, will enable the identification/detection of populations in specific geographies more likely to default on treatment and at-risk for selection of drug resistance.
Within the first year we anticipate to scale the solution to one country and expand this to 3 countries by year 3 by leveraging the existing SystemOne footprint. Long term, the platform could readily scale to the 40+ countries where Aspect is being deployed. The development of this solution would provide a new software module to the existing customer groups which can immediately be activated in the outbreak response in each country. Using environmental and publicly available data sources the Epidemic Control Platform can expand geographically as demand is being increased.
Furthermore, once the model is in place, adding new disease profiles using the high resolution real world data and evidence allows the model to identify vulnerable population groups (co-morbidities) and helps inform scientific partners about the progression of the disease within certain target audiences.
SystemOne engages with its Customers (MOH) via weekly meetings, management of support to a global Service Level Agreement, and continue to monitor system impact looking at a number of key (internal) metrics such as: Connectivity Uptime (%) - with a global target of 95% up-time for instrument communication, Percentage of positive diagnostic results notified, Time-to-Result Transmission, User Actioning of Result, Patient Initiation on Treatment.
EPCON measures the situation and program yield against the AI-steered program output through interactive near real-time dashboards and performance metrics. The model performance is constantly measured through retrospective validation, leave one out and linear regression models.
In addition, user and program meetings provide constant feedback about model output, relevance and user experience.
Integration of these two systems would be monitored as the time to detection of outbreaks and the efficacy of the response measuring the actual response against several possible forecasts. This would be measured by: Total number of test results provided compared with standard baseline offerings; Total number of patients initiated on therapy and associated outcomes compared with historical data; impact on morbidity and mortality including cost savings from limited exposure to the health system arising from faster turnaround time and treatment initiation.
- Azerbaijan
- Bangladesh
- Botswana
- Brazil
- Burkina Faso
- Cameroon
- Congo, Dem. Rep.
- Equatorial Guinea
- Ethiopia
- Ghana
- Guatemala
- Guinea
- India
- Indonesia
- Kenya
- Lesotho
- Liberia
- Malawi
- Moldova
- Mozambique
- Myanmar
- Nepal
- Netherlands
- Nigeria
- Pakistan
- Papua New Guinea
- Philippines
- Russian Federation,
- Sierra Leone
- South Africa
- Sudan
- Eswatini
- Tajikistan
- Tanzania
- Uganda
- Ukraine
- Vietnam
- Zimbabwe
- Mozambique
- Nigeria
- Pakistan
The primary barrier for both SystemOne and EPCON is funding. Both companies are relatively small organizations who do not have the operational budgets to support the implementation of new technologies without external funding. The resources supplied by this grant would enable the two organizations to focus time and effort on building the world’s first real-time surveillance platform leveraging laboratory diagnostic devices and AI/ML predictive modelling for disease and outbreak detection and response. SystemOne does not foresee additional barriers in terms of technical, legal or cultural barriers.
EPCON
Funding: rolling out the epidemic control platform for a given region is demand driven. Although the team is capable of driving the implementation of the system independently, the focus is limited to the regions with a specific ask.
Program data: publicly available data like demographics, land use features, air pollution, climate and others are readily available and pre-loaded on EPCON servers. Programmatic data however is often paper-based. Hence a collaboration with SystemOne, having a direct integration with molecular diagnostic systems can greatly accelerate the roll-out.
- For-profit, including B-Corp or similar models
SystemOne, LLC is the US-based, wholly-owned operating subsidiary of SystemOne Technology, Inc. and SystemOne SA PTY, LTD is the South African subsidiary. Customers include USAID,US-CDC,various national governments,NGOs.
EPCON BV, is a majority South African-owned Belgian company with participation of Imec research. Partners/customers: KIT Royal Tropical Institute,IHVN,National TB Program in Pakistan.
The reason for the application is to explore the acceptability of a health solution for disease control programs, that integrates rapid access to test results, instrument status, time to treatment initiation with AI/ML driven predictive analyses to optimize surveillance and outbreak mitigation in targeted geographies. Both SystemOne and EPCON are small companies operating in LMICs and global health. Neither entity generates sufficient revenue to be able to invest the necessary resources required to develop and scale this implementation without external funding. The Trinity Challenge uniquely targets infectious disease, data, and response to the data - which SystemOne and EPCON are uniquely positioned to respond to with real-time solutions and significant field experience and network within the various Ministries of Health.
N/A
Medical Scientist

Managing Director
Chief Operating Officer