Study Overview

Led by Dr. Mao Mao and a multidisciplinary team from leading Chinese hospitals, the multicenter retrospective study enrolled 1,814 participants, including 1,095 lung cancer patients. Blood samples were analyzed for four protein markers. The AI-driven algorithm combined protein marker levels with participant age and gender to detect lung cancer and its subtypes.

Two-Step Screening: A Game-Changer

To address the high false-positive rate and cost of low-dose computed tomography (LDCT), the researchers modeled a two-step approach. High-risk individuals are first screened with LungCanSeek; only those testing positive are referred for LDCT. This model not only drastically reduced unnecessary follow-up procedures and associated costs but also maintained a high detection rate of true cancer cases. The approach could enable population-wide lung cancer screening programs, even where access to expensive imaging is limited.