AI in medicine and care: Using opportunities responsibly
Artificial Intelligence (AI) can make an important contribution to improving medicine and care. Where the greatest potentials and hurdles lie for the use of AI in healthcare and what we need to do to use what is technically possible in a responsible way were discussed by AI experts at the third webtalk in the "AI.Impulse" series of Plattform Lernende Systeme. The event, which took place on 16 December at the Urania Berlin and virtually, was supported by Berlin Partner and the Cluster Gesundheitswirtschaft Berlin-Brandenburg.
AI assistance systems can help doctors diagnose diseases, predict complications in the individual course of a disease or support the decision for the best individual therapy. In nursing, intelligent speech recognition facilitates the nursing documentation of professionals and prevention systems based on AI can prevent patients from falling. All of these AI systems are based on large amounts of data that are evaluated in a short time with the help of machine learning. Medicine has always used data to make evidence-based diagnoses or choose therapies - whether analogue through microscopes or X-rays or digital through ultrasound or MRI, explained heart surgeon and computer scientist Alexander Meyer from the German Heart Centre Berlin. In the course of digitalisation, more and more data is available from the environment, from smart devices or about the patient himself. Today, the data sources are more diverse than ever, Meyer emphasised. AI is a powerful tool for using this data. In cardiac surgery intensive care units, an AI system could monitor ECG data, blood pressure and other vital signs of patients and provide early warning of post-operative complications even before the first symptoms appear.
Versatile applications: Personalised orthopaedics and follow-up care for heart surgery
If sufficient data is available on a patient, a digital twin can be created with the help of Artificial Intelligence, i.e. an exact virtual image of the person that enables individualised medical treatment. Güngör Kara, Chief Digital Officer of Ottobock and member of the steering committee of the Plattform Lernende Systeme, explained how customised prostheses can be manufactured and adapted in this way. In medical practice, AI assistance systems will be able to include global patient data and studies in their analyses in the near future and use this as a basis to provide doctors with recommendations for diagnoses and therapies. When advising on the best therapy, doctors can not only activate the knowledge of, for example, a tumour board in the room, but also the bundled world knowledge, says Klemens Budde, senior physician at the Charité Berlin and head of the working group "Health Care, Medical Technology, Care" of Plattform Lernende Systeme.
Learning to deal competently with probabilities
However, in order to exploit the potential of AI systems, medical and nursing professionals must be trained to use them. For example, AI requires dealing sensibly with probabilities, said Matthieu Schapranow, software engineer at the Digital Health Center of the Hasso Plattner Institute and member of Plattform Lernende Systeme. An AI assistance system does not offer one hundred percent certainty that its recommendation will deliver a correct diagnosis or the desired therapeutic success. Ultimately, the human must decide whether to follow the AI recommendation and which is the best choice for his or her patient. This must be reflected in training, Schapranow demanded in the discussion with the audience.
In addition to a competent handling of the systems in use, the experts underlined that AI applications must also be developed responsibly. The quality of the training data, the protection of sensitive patient data as well as international standards for data collection came up here.
The AI experts agreed on the role of humans in the medicine of the future: AI systems are helpful tools for routine activities and data evaluation, but cannot replace medical professionals. According to Schapranow, they lack human intuition and an eye for the particular.
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