Effective early screenings for gastrointestinal diseases are crucial for reducing mortality through timely interventions and improving life expectancy. In this paper, a strain effect-based biomimetic artificial whisker system is proposed to extract the structural and textural information of the tissues in the lumen based on interactive tactile perception data and an end-to-end screening algorithm. Benchmark experiment of the proposed method and an ex-vivo pilot study are conducted to characterize the baseline performance and feasibility of detecting several common tissue structures in surgical application scenarios. Our method shows promising results, with a test accuracy of up to 97.27% and a kappa value of 0.9590. This integrated hardware-software, end-to-end solution is promising to become an emerging human-machine interaction paradigm, empowering traditional healthcare applications.