Human-AI Framework to Investigate New Promising Oncological Radiotherapy Techniques

Abstract

In recent years, Artificial Intelligence (AI) has shown promise for accelerating discoveries and improving scientific advancement in the field of research on effective new cancer treatments. However, there is a lack of a framework that systematically exploits the potential of AI in research programs to study novel radiation therapy approaches, which require a strong interaction with human expertise. We propose a framework integrating AI into multiple stages of a research pipeline devoted to the clinical translation of new FLASH and mini-beam radiotherapy techniques, but flexible and adaptable to study other emerging radiotherapy modalities. We detail a workflow scheme highlighting the potential of AI in each key step, including advanced image analysis, multi-parameter correlation analysis, and predictive modeling of treatment outcomes. We also stress the importance of a comprehensive data platform to organize experimental data to foster AI-based analysis, considering the possibility of integrating AI tools into the platform to streamline research. Although AI enhances data processing speed and precision, the researcher’s knowledge of physics and biology remains crucial for contextualizing AI insights. We aim to stimulate the research community on how merging human expertise with AI capabilities could accelerate advanced radiotherapy techniques.

Ähnliche Arbeiten