Multidimensional decision support by artificial intelligence across the full workflow of endoscopic submucosal dissection for early gastric cancer.

Early diagnosis and treatment are essential for improving the prognosis of patients with gastric cancer. Endoscopic submucosal dissection (ESD) is the preferred minimally invasive treatment for early gastric cancer (EGC). Yet the entire diagnostic and therapeutic workflow depends heavily on the physician's experience and is subject to strong subjectivity, procedural variability, and a high rate of missed diagnoses. Artificial intelligence (AI) offers a new approach to addressing these clinical pain points. This review systematically examines the progress of multidimensional AI decision support across the full ESD workflow for EGC. It evaluates the applicable scenarios and limitations of distinct technical pathways, discusses the feasibility of cross-stage integration of AI systems, and analyses the core barriers to clinical translation, including the scarcity of multicentre standardised datasets, limited model interpretability, and poor integration with clinical workflows. Finally, the review anticipates future directions, such as multimodal data fusion and the combination of edge computing with augmented reality technologies. AI is driving ESD diagnosis and treatment towards greater precision and intelligence. The development of an intelligent clinical decision support system (CDSS) that integrates diagnosis, treatment and follow-up may further empower endoscopists to deliver individualised precision therapy for early gastric cancer.