Question
Answer
Where do the data come from?
The AI Health Shield™ relies on very large amounts of freely available public research data in medicine and health. Currently, about 30 different sources of data are used, including PubMed, patents and others. Although we estimate there are over 100 sources of biomedical and health research information, we have found only about 30 to be reliable in successive years and to be non-redundant.
In addition, the AI Health Shield™ is able to use EMR/EHR data in federated mode, as per published papers co-authored by the AI Health Shield™ team, the FDA and health insurance carriers. In federated mode, the EMR/EHR records are not accessed by the AI Health Shield™ at all. Instead, it builds the scenarios first and the owners of the EMR/EHR records run the specific math analytics based on the AI Health Shield™ scenarios.
How are the recommendations generated?
The inputs from the healthcare provider/patient are uploaded into the AI interface. Then, specifically designed workflows are followed that uncover credible connections between diseases, symptoms and possible medicines. Multiple “runs” need to take place before there is a set of insights that make sense and can be supported by evidence. Finally, human experts will always oversee the data and results before approving them for inclusion in the AI Health Shield™ deliverable.
What diseases are covered?
The AI Health Shield™ is unique because it covers almost every reported human medical condition and outcome. In this sense, it complements and goes beyond other tools, products or services that are focused primarily on cancer.
What medicines are covered?
The AI Health Shield™ examines every approved drug for its possible relevance against any healthcare condition that it receives.
How often is the AI Health Shield™ updated?
It is updated a minimum of three times per year with the latest data.
How should the recommendations be used?
The inputs from the healthcare provider/patient are uploaded into the AI interface. Then, specifically designed workflows are followed that uncover credible connections between diseases, symptoms and possible medicines. Multiple “runs” need to take place before there is a set of insights that make sense and can be supported by evidence. Finally, human experts will always oversee the data and results before approving them for inclusion in the AI Health Shield™ deliverable.
What’s a family subscription to the AI Health Shield™?
We make it possible to buy a subscription for a family, where up to three family members can each benefit from the AI Health Shield™ work. This complements the availability per single patient and makes particular sense for multi-generational families.
How is data privacy handled?
The AI Health Shield™ takes individual privacy extremely seriously. The data are not linked to individual patient identities. Individual patient cases are not “resold”.
Is there bias in how The AI Health Shield™ works?
There is no bias by disease or medication in the database that powers the AI Health Shield™ or in its analytics. This is by design, in order to make it possible to address all known human diseases and medicines with equal Augmented Intelligence “power.”