TL;DR
This paper uses advanced biomagnetometry and magnetic field tomography (MFT) to pinpoint the brain regions responsible for the contingent negative variation (CNV), a slow electrical potential observed during anticipation. They demonstrate that the CNV is generated by a distributed network of brain areas, not a single source, and that some of these areas, particularly the supplementary motor area (SMA) and posterior cingulate cortex (PCC), show remarkable plasticity with repeated task exposure.
What Problem Does This Paper Address?
The contingent negative variation (CNV) is a well-known electrophysiological signal that appears between a warning stimulus (S1) and an imperative stimulus (S2) in anticipation of an event. While its existence has been known for decades, the precise neuronal generators within the brain have remained elusive. Previous studies, often limited by the spatial resolution of their measurement techniques, have struggled to identify the specific brain regions contributing to the CNV, especially its early and late components. This ambiguity makes it difficult to fully understand the cognitive processes underlying anticipation, preparation, and response inhibition. This paper aims to resolve this by using a more comprehensive approach to localize the sources of the CNV with greater precision.
Methods
The researchers employed magnetic field tomography (MFT), a technique that reconstructs the distribution of neuronal current density within the brain from magnetoencephalography (MEG) data. They used two types of MEG systems:
- Experiment 1: A 2x37-channel biomagnetometer system (BTi MAGNES II) was used to record data from three healthy male subjects. This system provided coverage of specific brain regions, primarily the auditory and sensorimotor cortices.
- Experiment 2: For one subject, a 151-channel whole-head biomagnetometer system (CTF Omega 151a) was used. This provided a more comprehensive spatial coverage of the entire brain.
The MFT analysis involved:
- Low-pass filtering data to focus on the slow components of the CNV (0-20 Hz).
- Integrating the square of the current density vector over time and space to emphasize the strongest activations.
- Defining source spaces (regions where current density was allowed to be non-zero) that were either hemispheric or covered the entire head, depending on the MEG system used.
- Projecting reconstructed current densities onto individual subject MRIs for anatomical localization.
The experimental paradigm involved a Go/No-Go auditory choice reaction time task, with varying inter-stimulus intervals (ISIs) and repetitions of runs to investigate the stability and plasticity of the identified activations.
Datasets
The study used data from five healthy male volunteers. For two subjects, the data was discarded due to unusual noise. Therefore, detailed comparisons were restricted to three subjects. Subject JD was a highly trained participant who had undergone multiple CMV experiments previously, allowing for the investigation of plasticity. No new datasets were introduced.
Key Insights and Ideas
- Distributed Generators: The CNV is not generated by a single focal point but by a network of distributed brain areas. This explains why earlier studies struggled to pinpoint specific generators.
- Consistent Core Network: Despite intersubject variability, the auditory and sensorimotor cortices were consistently identified as key contributors to both early and late CNV components across subjects.
- Plasticity in SMA and PCC: The supplementary motor area (SMA) and posterior cingulate cortex (PCC), areas better covered by the whole-head system, showed significant lability. Repeated exposure to the same task led to a marked reduction in the amplitude of their activations, particularly in the GO condition, suggesting a form of neural plasticity.
- MFT's Power: Magnetic Field Tomography proved effective in identifying focal activations and their temporal sequencing, even in single subjects, without requiring prior assumptions about generator locations.
- Prefrontal Cortex Role: The study reinforces the known role of the prefrontal cortex, specifically the inferior prefrontal area (IPF), in motor inhibition and anticipation, showing consistent activation differentiating GO and NOGO conditions.
Key Results
- Auditory and Sensorimotor Cortex: These areas were consistently identified as sources of early and late CMV components in all three subjects.
- New Areas Identified (Whole-Head System): The whole-head system revealed activations in the Supplementary Motor Area (SMA) and Posterior Cingulate Cortex (PCC), which were likely missed by the more limited coverage of the twin-probe system.
- Lability of SMA and PCC: For subject JD, repeating the same experimental run led to a significant reduction in SMA and PCC activations, especially in the GO condition. This effect was observed even after extensive training.
- IPF Stability: The left inferior prefrontal area (IPF) showed stable and consistent activation, differentiating GO and NOGO conditions across runs.
- Early vs. Late Components: The early CMV component was observed in the auditory cortex, while the late component was more prominent in the sensorimotor cortex.
Limitations and Future Work
- Intersubject Variability: The study acknowledges large intersubject variability, which they suggest might reflect differences in strategy, training, or innate neural organization rather than just noise.
- Limited Subject Number: While the study used three subjects for Experiment 1 and one for Experiment 2, the findings, especially regarding plasticity, are primarily based on the detailed analysis of a single highly trained subject (JD). Generalizations require more extensive studies.
- Formidable Analysis: The authors note that the lengthy nature of the experiments and the complex analysis required for MFT make such undertakings "potentially very useful but also formidable."
Future work would involve replicating these findings with larger subject groups and exploring the mechanisms underlying the observed plasticity in areas like the SMA and PCC. Further investigation into the precise functional roles of these distributed generators in anticipation and response preparation is also warranted.
Summary updated