Journal: Advanced science (Weinheim, Baden-Wurttemberg, Germany)
This study presents an AI-powered SERS (Surface-Enhanced Raman Scattering) chip that enables early and accurate detection of ten common cancer types from a single serum test.
The platform uses a peptide-functionalized chip to selectively capture exosomes from patient serum, enhancing label-free Raman signals.
AI-driven analysis of these molecular fingerprints achieved high accuracy:
- 97.4% for distinguishing cancer from healthy samples,
- 97.08% for early-stage cancer detection, and
- 93.89% for classifying cancers including breast, thyroid, esophageal, kidney, pancreatic, duodenal, lung, colorectal, ovarian, and gastric cancers.
The study also identified exosomal deoxyadenosine triphosphate as a potential pan-cancer biomarker consistently elevated across multiple tumor types.
This fully automated, scalable system shows strong potential for clinical application in early and differential cancer diagnosis.