Each year, approximately 2.7 million people across the EU are diagnosed with cancer, with approximately 1.3 million deaths. A variety of factors, including genetic, lifestyle, and social elements, influence disease development, and the overall prevalence of cancer is expected to increase in the coming decades. Without action to prevent the disease, the number of newly diagnosed people is expected to increase to over 3.2 million by 2040.
The EU’s Innovation Mission on Cancer focuses on “improving the lives of more than 3 million people by 2030” through cancer prevention and cure, while helping “those affected by cancer including their families to live longer and better”. It aims to:
Increase understanding of cancer
Investigate where, how, and when cancer develops, identifying who is at higher risk of developing cancer and which treatments are more effective for different patients.
Enhance prevention and early detection
Promote awareness of the risk factors that contribute to cancer development, increase public and scientific understanding of those risks, and boost the effectiveness of screening programmes through better access, new methods, and early predictors.
Improve diagnosis and treatment
Increase the speed and effectiveness of diagnosis and treatment by sharing best practices, developing new methods, and delivering innovations in personalised medicine and clinical trials.
Increase quality of life for patients and their families
Integrate patient needs into treatment pathways, by analysing patient perspectives, enabling patient access to data, and building understanding of the impact of childhood cancers.
There are multiple points along the prevention, diagnosis, and treatment pathway where AI-enabled interventions could help deliver better health outcomes for cancer patients. For example:
Medical imaging for detection, diagnosis, and monitoring
Analysis of images such as X-rays or MRIs is central to many cancer screening and diagnosis methods. Leveraging the ability of AI to extract and analyse a large number of features from an image, a variety of AI tools have been developed to help increase the speed of diagnosis and help human clinicians determine the presence or absence of disease. Areas of application include analysis of lung radiography, CT scans, brain MRI scans, and mammograms, amongst others.
Precision oncology for diagnosis and treatment
Precision analysis of individual tumour cells can increase the effectiveness of cancer treatment, through targeted administration of oncology drugs (or other treatments) based on the genetic characteristics of a patient’s tumour. AI can enhance this single-cell analysis, helping to profile the genetics of cancer cells and to identify mutational signatures that help tailor treatment plans. By analysing how the proteins in cancer cells interact with each other and with different drugs, AI can help predict which cancer drugs are more likely to be effective for individual patients, helping tackle treatment resistance.
Clinical decision-support tools
By integrating different data types and extracting insights that might not be visible to human decision-makers, AI can contribute to new decision-support tools that help clinicians interrogate patient data and identify appropriate interventions, for example combining data from diagnostics with information captured in clinical reports. The natural language interfaces offered by Large Language Models offer a new route to increasing the usability of such tools. AI can also aid the development of robotics technologies used to support surgical interventions.
Cancer research
Underpinning many of these applications is the use of AI in cancer research, advancing understandings of cancer development and progression through analysis of genetic, environmental, lifestyle, and other types of data, and building risk profiles that interrogate the probability of different patients developing different diseases.
