Back to search

FRIMEDBIO-Fri prosj.st. med.,helse,biol

When default is not default: Solutions to the replication crisis and beyond

Alternative title: Når standard ikke er standard: Løsninger på replikasjonskrisen, og veien videre

Awarded: NOK 12.2 mill.

Functional magnetic resonance imaging (fMRI) is a widely used method for studying brain functions, both in healthy participants as well as patients. The development of fMRI more than 25 years ago was a milestone in brain research, and has lead to several major discoveries of how the brain functions. However, fMRI has one crucial limitation: it can almost exclusively be used for group studies. The reason for this limitation is that fMRI does not measure brain activity directly, only the metabolic processes assumed to be a consequence of brain activity, which is hugely variable between and within individuals and is easily affected by factors, like whether participants had a cup of coffee, have slept poorly the night before, had a stressful day, as well as a range of other, as of yet, unknown factors that could influence the measures. This variability of the fMRI signal makes it difficult for researchers to replicate earlier studies, and this problem is at the core of the ?replication crisis? that several researchers points to as one of the biggest challenges. However, there are several ways out of this crisis: First, one could assume that the underlying brain activations are much more reliable than the measured fMRI signal, a signal which only gives us an idea of how the brain activation may have looked like. Hence, the project will develop new methods that reliable can estimate the «true» brain activation based on the fMRI signal. Second, the «naturally» occurring variability of fMRI results will be examined. Perhaps there is no «replication crisis», but all diverging results are instead within the expected range the method affords? Once this is known, it will be possible to develop new methods of analysis that account for this «natural» variability. Lastly, once we reach a better understanding and new methods are developed, fMRI might be ready to take the next step: Broader clinical applications, like patient-focused and individualised diagnostic procedures. Since the start of the project, we have replicated across several open-access databases and own acquired data that the time of the day matters, when the fMRI data are acquired. Further, and as predicted, the metabolic character of the fMRI signal varies to a larger extent than the underlying neuronal activation. Further, the first results from one of our longitudinal studies, indicate that there is substantial variability across subjects, in how reliable the intra-individual results are. We expect that our subsequent analyses will identify critical parameters that influence this individual variability. Lastly, we are working on simulations and new analysis methods to provide new models of explanations.

This project addresses a very timely and highly important question of cognitive neuroscience studies: What factors influence the reliability of neuroimaging studies? Recent studies have estimated the reproducibility of psychological studies to be 39% or less and indicated a severe limitation of neuroimaging (fMRI) study reliability. Too small sample sizes, low to moderate effect sizes, and only partly understood neurophysiological mechanisms behind the BOLD/fMRI signal make it difficult to generalize results to larger populations, thereby impeding the impact of highly needed neuroscience studies on theoretical (scientific), methodological, and clinical progress. The overall objectives of this project are to (i) improve our understanding of the neurophysiological mechanism of the BOLD signal and its sources of variability to (ii) find a solution to the replication crisis by developing a new Bayesian and topology-based resting-state analysis framework as an alternative approach to today's analysis strategies, and to (iii) induce a paradigm shift from the current focus on an easy to measure but susceptible BOLD signal to the underlying, but (partly) hidden neuronal states that are presumably more stable and reliable. The project will break new ground by providing new insights into the neurophysiological mechanisms of the fMRI signal, its variability, dependency on endogenous and exogenous parameter, and reliability, and it will advance the research field of basic and clinical neuroimaging by providing new analysis strategies. To accomplish its mission, this project will use a longitudinal approach to examine the variability and reliability of resting-state networks and brain dynamics, supplemented by a cohort study in healthy participants as well as an exemplary clinical study and by the analysis of fMRI data from open-access data repositories.

Publications from Cristin

No publications found

No publications found

No publications found

Funding scheme:

FRIMEDBIO-Fri prosj.st. med.,helse,biol

Funding Sources