Interfacial Endoscopy: A Paradigm for Interface-Based Precision Diagnosis and Same-Access Intervention in Gastroenterology

For half a century, digestive endoscopy has advanced under a single dominant paradigm — morphological endoscopy — whose object is the morphological lesion, whose epistemic path is look–classify–confirm by pathology, and whose endpoint is resection of what is seen. That paradigm has delivered enormous value, but the clinical problems now dominating gastroenterology — antimicrobial resistance beyond the reach of empirical therapy and checkpoint-resistant, immunologically “cold” tumors — are problems of interfacial state rather than of form, and they mark the object boundary of a paradigm built on morphology. Meanwhile, the elements of a successor paradigm are already in the endoscopy suite — confocal microscopy, molecular imaging, artificial intelligence, in vivo pathology, same-access resection — yet none has simultaneously changed the object, the epistemic path, and the practical endpoint. We therefore propose interfacial endoscopy as a paradigm of equal standing: within a single endoscopic access, pathogen and immune phenotypes at the mucosal interface are read out in a one-stop manner, and the same access performs targeted in situ intervention — read the interface, classify the phenotype, intervene at once. The paradigm's three-layer structure (ontological, epistemological, praxical) is grounded in two published precursors from our program — the HP-REVISE closed loop for H. pylori pathogen phenotyping and ROSE-mIF immune phenotyping linked to same-access oncolytic intervention — and it stands before a defined five-item validation agenda. If validated, interfacial endoscopy would reposition gastroenterology as an interface-based discipline of in situ precision diagnosis and therapy, elevating the endoscope from the “eyes” of the field to its “eyes, readout, and hand”.