Fuse Presents at SROA on Unleashing AI in Radiation Oncology

At the Annual Meeting of the Society of Radiation Oncology Administrators in Washington D.C., Fuse leaders demystified terminology, concepts, and applications surrounding AI for attendees seeking to better understand how the tool will fit into their clinics. As generative AI continues to improve for more widespread use in rad onc, decision-makers need to know the implications for clinical workflow, patient care delivery, and general operations, particularly in light of staffing shortages, increasing time to treat, and reimbursement challenges.

Chief Technology Officer Christel Smith, PhD, led the discussion by addressing the current healthcare climate, pressing questions in the AI arena, and the presenters’ goals in providing answers. She also later explored the importance of utilizing ChatGPT with protective measures in a secure, HIPAA-compliant way, and safety stopgaps in document generation.

AI’s Role in Rad Onc

Data Scientist Joe Elnady, our resident AI expert, provided an AI primer with terminology explanations including rules-based systems, supervised and unsupervised machine learning, natural language processing, deep learning and neural networks, and large language models as well as relevant examples. He highlighted how AI is transforming clinical workflows, improving patient care, and addressing operational challenges like staffing shortages and increased time to treat.

Generative AI’s Impact

Director of Innovation and Senior Physicist David Wiant, PhD, discussed real-world applications of AI in healthcare, particularly its use in treatment planning and auto-contouring. He addressed the potential of generative AI in healthcare, including document generation, while emphasizing the importance of implementing AI in a secure, HIPAA-compliant manner.

Key Areas for Improvement

Vice President of Clinical Consulting Matthew Terry explored the data limitations further, addressing the importance of clean, consistent, and standardized data. He explained the shortcomings of OIS platforms and varying third-party systems that allow too much variation—leading to inconsistent data capture. AI tools can help correct these inconsistencies, improving workflow efficiency and decision-making consistency.

Looking Ahead

Christel concluded the Fuse SROA presentation with a summary of the bright outlook for future applications of AI given its rapid advancement. Implementing AI into clinics and workflows will be essential for rad onc success. It’s time to unleash its transformative power!