What we did:
For a molecular diagnostics client, combined high-throughput biological data with AI to identify cancer signals from liquid biopsies. Developed a deep learning model to uncover hidden patterns within complicated genomic, proteomic, and metabolomic datasets.
How it works:
Why it matters:
Detecting cancer signals in early-stage malignancies, where biomarker concentrations are extremely low, enabled our client to provide oncologists and patients with a comprehensive RUO report bolstering clinical evidence. We improved our client’s existing on-market testing accuracy by analyzing 20+ analytical targets simultaneously, providing more objective risk predictions compared to traditional single-threshold clinical methods.