AI-Driven Blood Test Detects Liver Cancer Across Diverse Populations
Johns Hopkins researchers validated an AI-powered DELFI liquid biopsy that detects liver cancer across two geographically and biologically distinct populations (Guatemala and Romania). When combined with AFP and basic risk factors, the test identified both early- and late-stage cancers with greater sensitivity than current blood tests, by analyzing genome-wide DNA fragment patterns that reflect signals from tumor, liver cells, blood vessels, and immune responses. The study also noted region-specific molecular signatures (e.g., aflatoxin-associated patterns in Guatemala) while the overall classifier remained effective across cancer causes. Future work will include prospective validation and integrating fragmentome analysis with other biomarkers for even earlier detection.
