Clinical Trials Navigator

The Clinical Trials Navigator intelligently matches patients to the most appropriate clinical trials, using advanced AI reasoning and LLM (Large Language Model) capabilities from Amazon Bedrock—integrated with knowledge stored in Neo4j’s graph database.

How the Clinical Trial Navigator works:

• Starlight’s query engine leverages data from the Holistic Patient Model, including detailed genomic test results, patient metadata (age, sex, clinical notes) and other diagnostic data
• User defines relevant clinical trials by filtering on key patient-specific criteria (age, sex, diagnosed diseases, co-morbidities)
• Starlight updates trial information from clinicaltrials.gov and NCI trial databases
• Starlight assesses the overall match of each trial with the patient profile using GenAI-driven ranking (maximizing inclusion potential and minimizing exclusion factors)
• Starlight evaluates the patient’s match with each inclusion and exclusion criteria for each trial
• Starlight identifies missing patient data needed to fully evaluate the trial (including the patient’s genomic profile and prior therapy regimens)
• The resulting trial assessments are displayed in descending order of likelihood to match for easier clinical review

This process seamlessly integrates patient data with clinical trial databases, providing accurate trial matching with Neo4j graph technology and advanced natural language processing tools.