AI methodologies have been applied to medical research for years, and have recently made an impactful entrance in oncology more specifically. AI broadly speaking consists in a set of techniques allowing computers to emulate human intelligence, employing algorithms created for the analyses and the design of either predictions or conclusions based on the analysis of big datasets.
The latter is especially important for cancer research considering the critical mass of data available for analysis and that standard analysis methods fail to exploit to its fullest potential. This is particularly the case for multiomics data, with their high variation in nature, format or storage.
The proper and effective and integration of these novel methodologies into the standards of clinical – but also basic and translational – research could prove to be an important leap forward for oncology research. Hence, this event will have two core training objectives.
The first will be to ease the clinical and research community into the field of AI methodologies themselves, still misunderstood or not known to its full potential – from a general overview of the most frequently used ML/DL methods and Explainable AI to a deep dive in novel data platforms and repository structures integrating these approaches in their design. This will allow clinicians to identify the value of AI models for their trials and studies, making the volume of patient – and tumor-related data valuable and more fully exploitable; as well as biologists to assess its potential in tumor biology to discover new biomarkers and mechanisms.
The second main endpoint will be to demonstrate not only the possibilities offered by the inclusion of AI models in standard practice, but really to present some concrete and innovative activities where they are already being successfully implemented. The focus is to demonstrate in particular the value of AI for both its predictive power and for the possibilities it opens up for the discovery of both new biomarkers and of new molecules targeting specific tumors. In particular, one section will be focused on the translational field and the synergy between AI-powered multiomic data analysis and clinical research, with regards to cancer immunotherapy, for example metastatic lung cancer.
In this context, we will showcase and discuss ongoing projects, including I3LUNG, along with other initiatives focused on the diagnostic and treatment pathways optimization for lung cancer patients. These projects involve collaboration among research institutes, academia, and private companies. As the AI research field is evolving at a rapid pace, the event will be topped off by a session offering perspectives already going beyond the current state of the art and providing insights into the Artificial Intelligence of tomorrow – how it could be involved as full-fledged actor in clinical decision-making.