Comprehensive Evidence Collection

First, Prognostics Engine uses probabilistic models for generating Patient Risk Profile based on patient information that we call Evidence. This is achieved by consulting with the Medical Knowledge-base which stores a large number of modeled medical conditions and interactions between them in the form of risk factors, shared symptoms and lab tests.

The engine calculates personalized probability of each modeled condition. The resulting Patient Risk profile contains pre-calculated probabilities, including a ranked list of conditions sorted by the personalized likelihood of having that condition and condition severity.

Intelligent Symptoms Triage

Second, the generated Patient Risk Profile is passed to the Diagnostic engine, which greatly increases the quality of the diagnosis and shortens the number of questions to ask.

The patient interview starts with the Chief Complaint. Then, the Diagnostic Engine dynamically creates an interview flow, where each next question is generated based on recalculating probabilities of hundreds of medical conditions, thousands of data points, and billions of relationships present in the Medical Knowledge-base.

Automated Differential Diagnosis

Finally, the Diagnostic Engine determines when to stop the symptoms triage, and delivers differential diagnosis in the form of most probable conditions that could've caused the symptoms along with more in-depth explanations for the diagnoses.