Stay in touch
Prime news from our network.
#read
Personalised medicine: experts still make better decisions than AI
For personalised cancer treatment, doctors need to analyse and interpret various data in a complex manner. Researchers are therefore investigating the extent to which artificial intelligence (AI) can take on this task in medicine. Their result: yes, it can, but not yet as well as human experts.
12/12/2023
If the body cannot repair certain gene mutations itself, this can lead to uncontrolled cell growth – a tumour develops. The decisive factor for this is an imbalance of growth-promoting and growth-inhibiting factors, for example through changes in oncogenes. Precision oncology, a specialist field of personalised medicine, makes use of this knowledge: overactive oncogenes are specifically switched off with the help of certain drugs such as small molecule inhibitors or antibodies.
In order for the doctors to know which gene mutations they can target with their treatment, they first analyse the tumour tissue genetically. Based on the molecular variants of the tumour DNA identified, they derive individual therapy recommendations.
In particularly complex cases, this requires knowledge from various medical fields. The so-called molecular tumour board (MTB) then comes together at the Charité in Berlin: Experts from pathology, molecular pathology, oncology, human genetics and bioinformatics jointly analyse the current study situation to determine which therapies promise the greatest success. A very complex process.Can AI identify therapy options?
Can artificial intelligence (AI) help here, asked Dr Damian Rieke, a doctor at Charité, Professor Ulf Leser and Xing David Wang from Humboldt-Universität zu Berlin and Dr Manuela Benary, a bioinformatician at Charité. Together with other researchers, they investigated the opportunities and limitations of large language models such as ChatGPT in the automated review of scientific literature for the selection of a personalised therapy.
„We set these models the task of identifying personalised therapy options for fictitious cancer patients and compared this with the recommendations of experts,
explains Damian Rieke. His conclusion: „Artificial intelligences were in principle able to identify personalised therapy options – but did not come close to the ability of human experts.“
Doctors versus AI in medicine
For the experiment, the team created ten molecular tumour profiles of fictitious patients. They then commissioned a specialist doctor and four large language models to determine a personalised therapy option. They presented these results to the members of the molecular tumour board for evaluation - without them knowing where a recommendation came from.
„Occasionally there were üsurprisingly good therapy options that the artificial intelligence identified“, reports Manuela Benary. „However, the performance of large language models is significantly worse than that of human experts.“ In addition, data protection and reproducibility would pose particular challenges when applying AI to real patients, according to Benary.
The best for medicine: AI plus expert
However, Damian Rieke is fundamentally optimistic about the potential applications of AI in medicine: "We were also able to show in the study that the performance of AI models continues to improve with newer models. This could mean that AI will be able to provide greater support in complex diagnostic and therapeutic processes in the future - as long as humans monitor the results of the AI and ultimately decide on therapies.
The above texts, or parts thereof, were automatically translated from the original language text using a translation system (DeepL API).
Despite careful machine processing, translation errors cannot be ruled out.