114 Artificial intelligence could soon play a crucial role in detecting hidden cancer during surgery, offering new hope for patients with advanced gastric cancer. Researchers have developed an AI system, known as the Artificial Intelligence Laparoscopic Exploration System (AiLES), that assists surgeons by identifying intra-abdominal metastases in real time during laparoscopic procedures. The findings were published in the journal npj digital medicine. Gastric cancer remains one of the deadliest forms of cancer worldwide, often diagnosed at an advanced stage when the disease has already spread within the abdomen. Accurately identifying metastasis is vital, as it determines whether patients can undergo curative surgery or must instead receive systemic treatments. Yet traditional methods, including CT scans and PET imaging, frequently miss small or hidden lesions, placing greater pressure on surgeons to detect them during exploration. The AiLES technology was developed using more than five thousand frames from surgical videos, covering a wide range of metastatic locations including the peritoneum, omentum, bowels, liver surface, mesentery and uterus. Its training dataset included images of tiny, single, multiple and extensive metastases, reflecting the diversity and complexity of cancer spread. Performance results were impressive. The system achieved a Dice score of 0.76, which measures the accuracy of identifying cancerous tissue, and demonstrated real-time recognition speeds of 11 frames per second. AiLES outperformed existing universal models like the Segment Anything Model (SAM), which struggled with the irregular shapes and textures common in surgical imaging. It also matched or exceeded the performance of novice surgeons in recognising challenging lesions, particularly tiny or occult ones that are often missed. Laparoscopic exploration remains a recommended diagnostic step for advanced gastric cancer, but success largely depends on the surgeon’s experience. Many younger surgeons, still building their skills, may inadvertently overlook minute cancerous implants. Missing these metastases can lead to incorrect staging, unnecessary surgery and a poorer prognosis for patients. AiLES offers a potential safety net, enhancing a surgeon’s ability to see and interpret hidden lesions without altering existing surgical workflows. Researchers suggest the system could eventually assist in calculating the peritoneal cancer index, an important tool used to plan individualised treatment strategies for patients. Although further development is needed to integrate features like lesion size measurement and anatomical region identification, the foundations laid by AiLES point towards a future where AI collaboration becomes routine in complex cancer surgeries. Despite promising early results, experts behind the project caution that larger, multi-centre studies are necessary to validate AiLES before widespread clinical adoption. The current dataset was drawn from a single centre and focused exclusively on gastric cancer, leaving open questions about how the system might perform across other types of abdominal cancers. Nevertheless, the real-time integration of AI into surgery marks a significant step towards improving accuracy and outcomes in one of the most difficult cancers to treat. By enhancing human vision in the operating room, AiLES could help ensure that no hidden tumour goes undetected.