Using AI to fight lung cancer – application scenario from Plattform Lernende Systeme

Cancer patients should be able to benefit more swiftly from research findings. This is the aim behind the National Decade Against Cancer, which was recently launched by the German Federal Ministry of Education and Research (BMBF). In its latest application scenario, Plattform Lernende Systeme demonstrates how Artificial Intelligence (AI) will be able to improve the recovery chances of cancer patients in the near future. The scenario illustrates how doctors can use AI-based assistance systems to access global medical references – from prevention to diagnosis and treatment – and thus boost the survival chances of lung cancer patients with the help of the latest findings.

Cancer is the second most common cause of death in Germany. There are nearly 500,000 new cancer diagnoses in Germany each year, and around 220,000 die from it. One of the most common forms of cancer is lung tumours, which only around one in five patients will survive beyond the five-year mark post-diagnosis.

At the heart of the “Fighting cancer with Artificial Intelligence” application scenario drawn up by the “Health Care, Medical Technology, Care” working group at Plattform Lernende Systeme is an AI-based assistance system that is set to help doctors with their decision-making in around five years’ time. Comprehensive medical expertise, the latest guidelines and worldwide, anonymised patient data will all be linked together on a data platform. The personal data of patients who consent will also flow into the platform. Artificial Intelligence makes it possible to assess and analyse the huge volumes of care data and the rapidly growing wealth of expertise. Doctors can use the results to provide tailored advice to their patients and improve the prevention, early diagnosis and treatment of lung cancer.

“We face the challenge of collating patient data from a range of different sources,” says Klemens Budde, Senior Physician at Charité Berlin hospital and head of the Plattform Lernende Systeme “Health Care, Medical Technology, Care” working group. “The information is currently stored in data silos, which means GPs, hospitals and health insurers only have access to their own records. Collating it all on a single platform requires the consent of patients to voluntarily hand over their information. This would enable us to use the data in anonymised format for research purposes or to improve the recovery chances of cancer patients.”

CT images of the lungs can also be analysed more precisely with the help of Artificial Intelligence. Neural networks will learn from several thousand precedences how to identify local deviations from healthy tissue. In operations, AI-based navigation systems will help surgeons by warning them when they get too close to a vital blood vessel, for example.

There are still some legal and technological challenges to overcome before cancer patients can benefit from self-learning systems, however. How can we protect personal data from abuse? Who is liable for any miscalculations made by the self-learning systems? And how can the assessments performed by AI systems be made transparent and understandable? The experts from science, business and society who make up the Plattform Lernende Systeme “Health Care, Medical Technology, Care” working group will address these questions in their recommendations for action, which are due to be published in mid-2019.

The application scenarios elaborated by the Plattform Lernende Systeme experts provide a glimpse ahead into the near future. They use selected topics to highlight what might be possible with the aid of Artificial Intelligence within just a few years, how humans stand to benefit, and which questions remain to be addressed. A second scenario demonstrating how self-learning systems can help emergency services respond to accidents and disasters is soon to be published.

You can find the oncology application scenario here (in German).

Further information:

Linda Treugut / Birgit Obermeier
Press and Public Relations

Lernende Systeme – Germany's Platform for Artificial Intelligence
Managing Office | c/o acatech
Karolinenplatz 4 | 80333 Munich

T.: +49 89/52 03 09-54 /-51
M.: +49 172/144 58-47 /-39
presse@plattform-lernende-systeme.de