Predictive brain signals best predict upcoming and not previous choices

Subjectivity as an Emergent Property of Information Processing by Neuronal Networks

Here, we examine subjectivity and consciousness as emergent properties of the computational complexity of information processing by the brain, rather than metaphysical phenomena. While Psychology concentrates on the emergent properties and Neurobiology examines the properties of the biological substrate, Neurophysiology and Cognitive Neuroscience link the two levels by investigating the mechanisms and processes by which the functions of the brain emerge from the anatomical, cellular and network properties of the nervous system. Our purpose here is not to locate the neural structures that sustain subjectivity or other psychic functions; rather, we examine the operating modes of neurons and neural circuits: they reveal an intrinsically relational quality; sensory elaboration itself proves to be relational and self-centred, necessarily associated with the vital, hedonic, emotional relevance of each experience and external cue, and intrinsically oriented to a behavioral interaction with the latter. The hippocampus adds to this self-centred relational perspective the capability of transforming the identification and the spatial location of objects into a contextualized representation of reality. Since the hippocampus is strongly interconnected with the archaic structures that evaluate vital and hedonic relevance and generate emotional responses, the contextualized information, emotionally colored, is transformed into a comprehensive individual experience. This way, a subjective, self-centred, affectively colored perspective arises in animals due to the intrinsic properties of neuronal circuits in the brain. We conclude that neuronal network processing is strongly characterized per se by a relational and self-centred (subjective) and emotionally colored, motivationally oriented (personal) perspective. The properties and features of neural processing discussed here constitute well-established knowledge in the neuroscientific community. Yet, from a layman’s perception, subjectivity still mysteriously arises in our brain due to the action of consciousness, and in epistemological and philosophical debates, the question often arises as to how consciousness may add the subjective and personal perspective to neural elaboration. The answer appears to be simple: it does not; subjectivity is already there, present ab initio in neuronal processing and not added a posteriori by some other “consciousness” function of unclear neural basis.

Decoding the contents and strength of imagery before volitional engagement

Is it possible to predict the freely chosen content of voluntary imagery from prior neural signals? Here we show that the content and strength of future voluntary imagery can be decoded from activity patterns in visual and frontal areas well before participants engage in voluntary imagery. Participants freely chose which of two images to imagine. Using functional magnetic resonance (fMRI) and multi-voxel pattern analysis, we decoded imagery content as far as 11 seconds before the voluntary decision, in visual, frontal and subcortical areas. Decoding in visual areas in addition to perception-imagery generalization suggested that predictive patterns correspond to visual representations. Importantly, activity patterns in the primary visual cortex (V1) from before the decision, predicted future imagery vividness. Our results suggest that the contents and strength of mental imagery are influenced by sensory-like neural representations that emerge spontaneously before volition.

Proactive Recruitment of Frontoparietal and Salience Networks for Voluntary Decisions

There is evidence that neural patterns are predictive of voluntary decisions, but findings come from paradigms that have typically required participants to make arbitrary choices decisions in highly abstract experimental tasks. It remains to be seen whether proactive neural activity reflects upcoming choices for individuals performing decisions in more complex, dynamic, scenarios. In this functional Magnetic Resonance Imaging (fMRI) study, we investigated proactive neural activity for voluntary decisions compared with instructed decisions in a virtual environment, which more closely mimicked a real-world decision. Using partial least squares (PLS) analysis, we found that the frontoparietal and salience networks were associated with voluntary choice selection from a time at which decisions were abstract and preceded external stimuli. Using multi-voxel pattern analysis (MVPA), we showed that participants’ choices, which were decodable from motor and visual cortices, could be predicted with lower accuracy for voluntary decisions than for instructed decisions. This corresponded to eye-tracking data showing that participants made a greater number of fixations to alternative options during voluntary choices, which might have resulted in less stable choice representations. These findings suggest that voluntary decisions engage proactive choice selection, and that upcoming choices are encoded in neural representations even while individuals continue to consider their options in the environment.

The same analysis approach: Practical protection against the pitfalls of novel neuroimaging analysis methods

Standard neuroimaging data analysis based on traditional principles of experimental design, modelling, and statistical inference is increasingly complemented by novel analysis methods, driven e.g. by machine learning methods. While these novel approaches provide new insights into neuroimaging data, they often have unexpected properties, generating a growing literature on possible pitfalls. We propose to meet this challenge by adopting a habit of systematic testing of experimental design, analysis procedures, and statistical inference. Specifically, we suggest to apply the analysis method used for experimental data also to aspects of the experimental design, simulated confounds, simulated null data, and control data. We stress the importance of keeping the analysis method the same in main and test analyses, because only this way possible confounds and unexpected properties can be reliably detected and avoided. We describe and discuss this Same Analysis Approach in detail, and demonstrate it in two worked examples using multivariate decoding. With these examples, we reveal two sources of error: A mismatch between counterbalancing (crossover designs) and cross-validation which leads to systematic below-chance accuracies, and linear decoding of a nonlinear effect, a difference in variance.