Our inboxes and social feeds have been flooded with news stories related to how AI is being used to help solve the coronavirus.
AI To Expedite Drug Discovery, Simulate Impact of Policy Decisions
We’ve highlighted some of our favorite below and are also excited to share Lumina’s capabilities to help predict the spread and identify those most at risk of exposure.
- University of Virginia researchers are using AI to simulate the impact of policy decisions on health outcomes in a particular community;
- Our local Tampa General Hospital is enlisting AI to help detect feverish visitors with a simple facial scan;
- The Federal government announced a collaborative effort to discover possible treatments for COVID-19 by using AI to mine the 2000+ research papers written on virus; and
- AI-based drug discovery platforms are working to predict which existing drugs, or brand-new drug-like molecules, could treat the virus.
Lumina’s AI-Driven Radiance can Help Triage Individuals – Even Before the Emergency Room
Here at Lumina, our Radiance platform and S4 crowd sourcing app provide federal, state and local governments, public health officials, and the healthcare industry tools to predict and better triage those individuals at higher risk for coronavirus – even before they reach the hospital or emergency room.
Radiance is a deep-web listening tool that uses machine learning and artificial intelligence to assess and prioritize risk. Radiance can also ingest large internal unstructured data sets and use its AI capabilities to assess and prioritize risk from that information.
Additionally, Lumina’s S4 (See Something, Say Something) app is another tool to help public health officials efficiently gather data about those who are showing symptoms of the coronavirus or have tested positive. Developed as a mobile app to help report suspicious behaviors, it has been updated to help crowd-source data about the coronavirus.
Crowdsourcing Coronavirus Symptoms and Cases
Armed with the S4 mobile app, frontline healthcare providers can quickly and efficiently report cases to the appropriate regulatory authorities. This real-time information creates an important data set for analysis and prediction of virus spread.
The app can also be made available on Google Play and the App store, and downloaded by the general public as well. Citizens can self-report symptoms or if they have tested positive, with that information made immediately available to the appropriate regulatory authorities.
The app links to the CDC website with information on what an individual should do if they think they have been infected.
Information reported through S4 is geofenced by hospital facility, and all reports from citizens are geolocated. The information is ingested into Radiance, which uses AI to analyze the data to show prevalence of exposure in a given area and help predict where the virus is going. The data is reflected in a heatmap to demonstrate spread of the virus in real time.
Radiance Predicts Potential Exposure with Coronavirus Behavioral Affinity Models
The names of those individuals being screened for potential coronavirus exposure, including those reported through the S4 app, are entered into Radiance and correlated with our proprietary Coronavirus Bundle, which includes two Behavioral Affinity Models (BAMs), encompassing more than 50 phrases, across 14 languages – all related to coronavirus risk.
- The Coronavirus Symptom BAM: This BAM consists of terms and phrases that may indicate an individual has been diagnosed or shows signs consistent with a positive case of COVID-19.
- The Coronavirus Hotbed BAM: This BAM consists of terms and phrases that are intended to identify potential new epicenters of the coronavirus outbreak. It uses known locations of high risk to help screen for individuals that have recently traveled to or are connected to those areas.
Internal Datasets Add Another Layer of Risk Assessment
Radiance can also ingest internal data sets, such as state health records, emails, human resources data or other pertinent information. Using the Coronavirus BAM, Radiance can cleanse the internal documents, creating an ecosystem of information relevant to coronavirus risks.
Proprietary Name Extraction Relationship Maps Those at Risk
Radiance’s proprietary name extraction capabilities provide another tool for analysis and relationship mapping.
Radiance can ingest open-source data ecosystem of URLs and text surrounding individuals in the United States who are known to have contracted COVID-19. This would include names of individuals reported through the S4 app. Radiance runs its proprietary name extraction process from these URLs to determine individuals who are digitally adjacent to those infected.
Continuous Monitoring and Delta Reporting
This process will run continuously, with the ability to monitor new names and relationships that enter the ecosystem as the virus progresses. These efforts can be extended to global data ingestion and relationship mapping.
The Needle in the Haystack
Keep in mind, Radiance’s deep web and internal data set search capabilities are designed to find the needle in the haystack.
Radiance’s power is in identifying all the content that includes the name being searched and a term related to the behavioral affinities. The fact that Radiance flags an individual as relevant does not mean that a risk is necessarily involved. The results are designed as a tool to help expedite a healthcare professional’s ability to assess risk beyond visibly obvious symptoms and prioritize patients accordingly.
Contact Us to learn more about the Radiance solution.