A new study from the Technical University of Munich reveals that up to 40% of signals in functional magnetic resonance imaging (fMRI) technology don’t actually correspond to real brain activity. This finding has significant implications for neuroscience research and medical diagnosis, potentially requiring reevaluation of past research conclusions based on fMRI. Researchers discovered that fMRI signals are influenced by multiple non-neural factors, including physiological processes like head movement, heartbeat, and breathing. This discovery not only challenges current understanding of brain activity monitoring but also presents new challenges for artificial intelligence model training that relies on fMRI data. Experts indicate that these research findings will drive the development of more precise brain imaging technologies, helping scientists more accurately interpret brain activity.
