72 A new artificial intelligence tool developed by researchers at La Trobe University could help predict which patients with stage two bowel cancer are at risk of relapse, potentially allowing thousands to receive life-saving treatment earlier. Published in the journal Gastroenterology, the study describes the creation of SÉMIL (Semantically Enhanced Multiple Instance Learning). The AI algorithm combines images and written descriptions to analyse routine pathology slides and identify patients at greater risk of recurrent bowel cancer. Researchers analysed more than 1,600 pathology slides and validated the findings in 1,220 patients with stage two bowel cancer across three independent cohorts from multiple Australian institutions. The study also found that combining AI-based assessment with pathologist evaluation produced the most accurate risk predictions. Lead author and PhD candidate Francis Magisson, from La Trobe University’s School of Computing, Engineering and Mathematical Sciences, said the technology assessed the tumour’s growth pattern at its invasive front, a feature that is important for predicting outcomes but difficult for pathologists to classify consistently. It then assigned each tumour to either a high-risk or low-risk group. “This information could be used to assist pathologists and clinicians in identifying which patients with stage two bowel cancer are at higher risk of relapse and may need closer monitoring or additional treatment, such as chemotherapy,” Magisson said. Current Australian clinical guidelines recommend chemotherapy after surgery only for patients with high-risk stage two bowel cancer, making accurate risk assessment particularly valuable. Associate Professor Zhen He, who leads the Digital Biology Program at La Trobe University, said the findings showed AI’s growing role in precision medicine. “SÉMIL could be integrated into existing digital pathology workflows without requiring expensive new tests or additional tissue samples, providing more consistent and objective assessments of cancer pathology,” Professor He said. “In future, AI-based pathology assessments could be combined with other emerging biomarkers to improve risk stratification and treatment planning.” Associate Professor David Williams, an anatomical pathologist from Austin Health and the Olivia Newton-John Cancer Research Institute, which is affiliated with the La Trobe School of Cancer Medicine, said the tool was designed to support rather than replace clinical decision-making. “One of the biggest challenges in stage two bowel cancer is identifying which higher-risk patients require treatment and weighing the potential benefits of chemotherapy against its side effects,” Associate Professor Williams said. “Our study shows that AI-based assessment has the potential to provide pathologists with an additional layer of information that could help clinicians make more informed decisions about subsequent treatment options.” Bowel cancer is Australia’s fourth most commonly diagnosed cancer and the second leading cause of cancer-related deaths. Worldwide, it is the third most commonly diagnosed cancer. Around 15,000 new cases are diagnosed in Australia each year, and more than 5,000 people die from the disease annually. Researchers said early and accurate risk assessment remains critical for patients with stage two bowel cancer, where decisions about additional treatment, such as chemotherapy, are finely balanced. The study involved collaborators from the La Trobe Centre for Molecular Science, the La Trobe School of Cancer Medicine, WEHI, Monash University, the University of Melbourne, UNSW Sydney, the Peter MacCallum Cancer Centre, and other cancer centres across Australia. Researchers said the AI system uses routine pathology slides that are already collected as part of standard care, avoiding the need for additional invasive procedures or costly molecular tests. This could make the technology easier to adopt across healthcare systems. They emphasised that it is intended to support multidisciplinary clinical teams, with final treatment decisions remaining in the hands of clinicians and pathologists. Further validation and prospective clinical trials will be important before the technology can be adopted more widely. Researchers said the work contributes to the broader field of AI-driven cancer diagnostics and could help improve the precision of bowel cancer care.