Science/Tech
Study Reveals Brain Scans Aren't Always Accurate
Brain scans, particularly fMRI, are widely used to map brain connectivity, but a new study reveals they can produce misleading results.
Impact of Arousal Levels on Brain Scans
Functional magnetic resonance imaging (fMRI) scans serve as non-invasive instruments designed to chart brain connectivity. They are crucial in planning surgeries, understanding stroke impacts, and studying mental illness effects on neurological functions. However, these brain scans may not always provide accurate results.
The study, conducted by experts from McLean Hospital, Harvard Medical School, and the National Institute on Drug Abuse (NIDA-IRP), highlights that changes in arousal levels during the scans can distort the results.
Physiological Noise
The primary issue identified in the study is the change in subjects' arousal levels during the scans. As people become more relaxed and sleepy, their breathing and heart rates change. These changes alter blood oxygen levels in the brain, which fMRI scans detect.
Unfortunately, these alterations are falsely interpreted as normal neuronal activity. Cole Korponay, a Research Fellow at McLean Hospital Imaging Center, explains that these conditions create an illusion of inflated brain connection strengths throughout the scan.
Identifying the sLFO Signal
The study identified a specific blood flow signal, known as the "systemic low-frequency oscillation" (sLFO) signal, which tracks the decline in arousal levels and the illusory inflation of brain connection strengths. This non-neuronal physiological noise grows over time during scanning and closely matches the pattern of increased connection strengths.
Implications for Future Research
Understanding and addressing the distortive effects of arousal changes is crucial for enhancing the accuracy of fMRI scans. The study suggests adopting the sLFO denoising procedure to mitigate these effects. This approach aims to improve the validity and reliability of fMRI findings by reducing the influence of physiological noise on the data.
Join the Conversation