The Role of AI in Early Breast Cancer Treatment

During the 2024 SABCS, Dr Barbara Bussels, radiation oncologist at the AZ Delta hospital in Roeselare presented two posters evaluating the added value of artificial intelligence (AI) driven clinical decision support systems (CDSS) in early breast cancer. In this video Dr Bussels and Ms Sandra Steyaert, data scientist at the RADar learning and innovation centre of the AZ Delta, summarize the key take aways from these posters. Within the context of the BreaCS consortium, clinical, pathology and radiology data of early breast cancer patients treated in 4 Flemish hospitals are being collected. Based on the data of approximately 5000 patients, three AI models were developed to pre-operatively predict the pT stage, the complete response rate after neoadjuvant chemotherapy, or the involvement of sentinel lymph nodes. According to the first of these posters, the use of these models would reduce the surgical costs for early breast cancer treatment with approximately 150,000 euro assuming a total of 1200 patients/year. The second poster demonstrated the potential of an AI CDSS model to predict the response to primary systemic therapy (PST) in early breast cancer, with enhanced predictive accuracy if a multimodal approach was used.2