Kahun’s ‘secret sauce’ is our ability to instantly bridge the gap between text-based research publications and clinical decision making.
The knowledge graph captures the complex inter-activities between thousands of diseases and findings. Our proprietary AI engine generates and infers from complex variants of large Bayesian networks that link diseases, findings, complications, risk factors, causal information, and statistics—all based on the most up-to-date medical information.
A mapped version of evidence based medical literature
By merging AI methodologies and a community of medical experts, we annotate and process existing clinical literature and fully capture its complex associations.
We encode the evidence-based clinical knowledge and transform it into a machine-readable medical knowledge graph.
The knowledge graph displays causal, associative and quantitative information about risk factors, clinical manifestations, lab abnormalities, imaging findings, and more.
Using standard terminologies such as SNOMED-CT and LOINC, Kahun speaks the same language as physicians and EHR systems.
Dynamic and growing daily with each new paper published, the knowledge graph keeps you current on millions of relations between disorders, complications, findings, incidence and prevalence rates, drug side effects, and more.
Our AI inference engine connects pieces of medical knowledge to your patient’s clinical presentation.
Generate a ranked list of differential diagnoses
Suggest options for further workup
Assess the factors that can impact the ranking
Evidence-based clinical interview
The intake is conducted with the same reasoning that should guide an experienced physician
A clear path to the knowledge
Each question is backed by a reasoning path that leads to the originating source of evidence
Tailored to your patient
Our AI algorithm considers millions of complex associations linking thousands of diseases and findings to come up with the next best question for your specific patient
AI that thinks like you
Hypothesis-driven data collection, not based on big data or static decision trees