The CD Lab Machine Learning Driven Precision Imaging, funded by the Christian Doppler Gesellschaft, will develop new predictive models for lung cancer and its individualized treatment.
It will integrate radiological and pathological images and molecular data using new machine learning methods. This paves the way for the development of new machine learning (ML) concepts in precision imaging for better individualized treatments.
The CD Lab aims at fighting primary lung cancer, one of the most common types of cancer and the leading cause of cancer death worldwide, more effectively. It is an interdisciplinary project across the fields of machine learning, medical imaging, oncology and pathology work, and legal research. Together with a team involved in practical legal issues and policy the CD Lab will develop and validate novel ML methodology to improve the individualized care of lung cancer patients.
This integration of AI (artificial intelligence) and imaging offers numerous challenges: Available routine clinical data are heterogeneous, the patient population is diverse, training data usable for AI models are limited in number, and keeping said models permanently updated to work with the availability of new therapies and the simultaneous development of imaging technologies is a highly complex endeavor, while the legal conditions for sharing data collections are yet challenging.
The CD Lab has 4 areas of research: First, to quantitatively assess and predict disease and treatment trajectories; second, to extend ML to the large, diverse, heterogeneous routine patient population, as well as to continuously evolving and emerging diagnostic technologies and treatment options, rather than assuming only focused studies. Third, to link evidence in the form of large-scale data on the one hand using ML to underlying biological processes, and finally, fourth, to clarify the legal requirements of data sharing, AI development, and use of AI in healthcare.