98 A new artificial intelligence system is revolutionising the way surgeons detect intra-abdominal metastasis (IAM) during laparoscopic gastric cancer surgery. The AI-assisted laparoscopic exploration system, known as AiLES, has demonstrated its ability to enhance tumour staging and improve the accuracy of treatment planning, potentially changing the landscape of gastric cancer management. The findings were published in the NPJ Digital Medicine. Gastric cancer remains one of the leading causes of cancer-related deaths worldwide, with early detection and precise tumour staging playing a crucial role in patient outcomes. Traditionally, IAM detection relies on laparoscopic exploration, but the process is susceptible to human error, particularly in identifying small, subtle lesions. The limitations of conventional methods often lead to understaging, which can affect the choice of treatment and long-term survival rates. The AiLES system was developed using a dataset comprising thousands of frames extracted from laparoscopic videos. By leveraging deep learning techniques, the AI model was trained to recognise IAM lesions with a high degree of accuracy. During testing, AiLES achieved a Dice score of 0.76 and demonstrated a real-time recognition speed of 11 frames per second. This performance places it on par with novice surgeons while surpassing them in detecting smaller and more occult lesions that are easily overlooked. One of the key advantages of AiLES is its ability to assist in intraoperative decision-making by providing real-time feedback to surgeons. The system integrates seamlessly into laparoscopic procedures, identifying potential metastatic lesions and ensuring that no critical details are missed. By reducing the likelihood of understaging, AiLES helps clinicians make more informed decisions regarding treatment strategies, whether opting for surgical resection or systemic therapies. The study behind AiLES compared its effectiveness with existing AI models and found that it outperformed several alternatives in detecting IAM. The AI system was particularly adept at recognising lesions across different metastatic locations, including the peritoneum, omentum, bowels, mesentery, and liver surface. Furthermore, it maintained strong accuracy across varying extents of metastasis, reinforcing its reliability in clinical settings. Despite its promising capabilities, AiLES is not without limitations. The current model is based on data from a single medical centre, which raises concerns about its generalisability to broader populations. Researchers acknowledge the need for further validation through multicenter studies and adaptation to different laparoscopic devices. Additionally, expanding the dataset to include more cases of rare metastatic patterns could enhance the system’s robustness. Experts believe that integrating AI into laparoscopic surgery represents a significant step towards precision oncology. By augmenting human expertise with machine learning, AiLES has the potential to refine surgical workflows and ultimately improve patient survival rates. Future advancements in AI-driven medical imaging and real-time surgical guidance may further enhance the accuracy of tumour detection, paving the way for more personalised and effective cancer treatments.