▎ 摘　　要

This paper proposes pattern recognition of gastric cancer biomarkers CEA, CA19-9 and p53 in whole blood and urine samples using a stochastic sensor based on exfoliated graphene (E-NGr) paste modified with protoporphyrin IX. The proposed sensor covered large ranges of concentrations: 1 x 10(-12)-1 x 10(-7) mu g/mL for CEA, 1 x 10(-13)-1 x 10(2) U/mL for CA19-9, and 0.2-5.0 mu g/mL for p53. These ranges allowed the determination of the three biomarkers from early to latest stages of gastric cancer. Validation of the pattern recognition of gastric cancer biomarkers was accomplished using biological samples: whole blood and urine.