Publications / Poster

Improving early cancer detection by training scalable deep neural networks to extract tumor signal from cell-free DNA

American Association for Cancer Research (AACR) Annual Meeting 2026

ProbeCDN™ is a deep-learning framework that detects cancer by analyzing methylation patterns at the level of individual DNA reads, overcoming limitations of traditional methods that dilute early tumor signals. With significantly improved early-stage cancer detection sensitivity, this framework marks an important advance in more accurate, non-invasive cancer detection.

Read More