Flashcard library · Neuroscience · PaperFren
Tracking Decision Making in the Brain Using fMRI
While many brain areas collect sensory evidence, only a specific network in the frontal cortex marks the exact moment a decision is made.
Keep what you just learned
Save these 35 cards to your NoteFren library. Your first review starts right away, and each card links back to the study it came from.
What is the core premise of the accumulation model in perceptual decision making?
Sensory evidence is accumulated over time for each alternative until the evidence for one option reaches a threshold, triggering a choice.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI
Where have neurophysiological studies in monkeys identified accumulator-like neural activity?
In the lateral intraparietal area (LIP), the frontal eye field (FEF), and the superior colliculus.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How did Ploran et al. (2007) classify decision-related brain regions using the temporal-profile approach?
They classified them into three groups: sensory processors, accumulators, and moment-of-recognition regions.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What methodology did this study use to recover approximate original neural activity from slow BOLD signals?
The fMRI deconvolution method, specifically Wiener deconvolution, was used to reconstruct the temporal shape of underlying neural activity.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What is the main finding of the study regarding how multiple brain regions participate in decision making?
Perceptual decision making is carried out by a distributed system of multiple functionally-heterogeneous regions, but only a subset (primarily in the frontal cortex) show peak activity matching the actual decision moment.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How was the sample size of 14 participants determined and justified?
Since standard power analysis isn't built for cluster analysis, an a priori power analysis was conducted on pilot data looking at standard deviations of peak times across RT bins between paired brain regions.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What three categories of stimuli were used during the decision-making section?
Human (with head and body parts), scene (architectural scenery), and tool (handheld tools).
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How were the degraded stimulus images produced for the gradual unmasking technique?
By applying 15 levels of spatial low-pass Fourier filtering, manipulating the high cutoff frequency from 1.0 to 4.5 cycles per degree.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How was the gradual unmasking temporal progression structured in each trial of the decision section?
The 15 progressively less degraded versions of an image were sequentially presented every 1.4 seconds.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What were the five stimulus categories used in the localizer section?
Face, body, human, scene, and tool.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why was slice timing correction omitted from the temporal-profile analysis of the decision section?
To preserve exact, unmodified signal values and instead model the precise acquisition times of each slice during averaging.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why did the study use the ICBM template of East Asian brains for spatial normalization?
Because all 14 participants in the study were native Asian (Japanese) individuals.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What was the purpose of the 12-second rest epoch following the stimulus epoch in each trial?
It provided enough time for the elevated BOLD signal to return to baseline levels before the next trial began.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How was prior knowledge manipulated in the decision task, and what did the authors expect?
Cue words (congruent, incongruent, or neutral) were presented before the stimulus, expecting that congruent cues would speed up decision/accumulation and incongruent cues would slow it.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why were cue conditions pooled together in the final fMRI analysis?
Cue effects were complex, stimulus-dependent, and participants noticed the high rate of incongruent cues early on, which led many to ignore them.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What trials were excluded from the fMRI timeseries analysis to prevent mixing multiple decision processes?
Trials where the participant's first button press was incorrect were excluded, keeping only correct-first-response trials.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How did the double-click verification instructions impact the activity profiles?
The impact was minimal, as verification responses were widely and almost uniformly distributed throughout the post-decision interval.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What is the 'individual peaks within group activation' method of ROI determination?
It identifies the voxel with the strongest task effect within each individual participant, but restricts this search to a spatial mask defined by group-level positive activation.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why was the search for ROIs restricted to regions showing positive activation during the task?
Because accumulator regions are expected to show positive, growing activity as evidence is gathered during a decision process.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How were Non-Category Selective Regions (NCSRs) mathematically isolated?
The individual-level category-selective maps for face, body, scene, and tool were summed and inverted to form a mask of non-selectivity, which was then intersected with the positive activation mask.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What atlas was used to divide the non-category-selective areas of the brain into anatomical structures?
The Tissue Probability Map (TPM) atlas available in SPM12, using its volume data for 121 cortical and nuclear regions.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why is direct comparison of deconvolved signal amplitudes across different brain regions limited?
The absolute magnitude of the hemodynamic response function (HRF) is arbitrary, which prevents comparing absolute amplitudes across regions.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI
How did the authors handle high-frequency noise when applying Wiener deconvolution?
They assumed a constant spectrum of white noise and estimated its energy from the highest frequencies of the grand-averaged BOLD signal power spectrum.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What mathematical solution did the authors use to prevent Wiener filter values from blowing up at high frequencies?
They added a positive term (delta = 24) to the denominator of the filter formula, raising its minimum above zero.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What systematic distortion occurs in the rise time of signals deconvolved using a Wiener filter?
Deconvolved signals rise slightly earlier than the true model neural activity due to the low-frequency transparency of the filter.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why did the study use two alignments (stimulus-locked and response-locked) to average trial segments?
To capture precise timing properties of neural activity relative to both stimulus onset and the varying moments of choice.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What three profile parameters were extracted to characterize the deconvolved temporal profiles of the brain regions?
Peak time of the stimulus-locked course (Peak_stm), peak time of the response-locked course (Peak_rsp), and rising slope of the response-locked course (Slope_rsp).
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How was clustering performed to classify the regions of interest into functional groups?
Hierarchical cluster analysis using Ward's method was applied in a 4-dimensional space based on the standard deviations and means of the profile parameters over RTs.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why was the rising slope (Slope_rsp) used for classification instead of the simple rise time?
The authors found heuristically that the rising slope was more effective for reproducing previous three-class classification results.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How did behavioral accuracy relate to reaction time (RT) in the category-decision task?
Accuracy increased with longer reaction times, which is a classic behavior predicted by accumulation models with varying thresholds.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Which cortical regions were identified as Category-Selective Regions (CSRs) in this study?
Two face-selective, two body-selective, three scene-selective, and one tool-selective regions (including equivalents of FFA, OFA, EBA, PPA, and LOC).
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How did the Category-Selective Regions (CSRs) behave during the degraded stimulus decision task?
Almost all CSRs maintained highly significant selective activation for their preferred stimulus category, even when the images were severely degraded.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
Why was the left scene-selective region 2 (L Scene-SR 2) excluded from the final CSR analyses?
It showed inconsistent category selectivity during the decision task, likely because its initial activation was driven by high spatial frequency components (lines/edges) in the uncropped scene backgrounds of the localizer task.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
What does the 'accumulation system' concept imply about the functional architecture of decision making?
Decision-making is not localized to a single brain region; it is handled by a distributed network of functionally diverse areas, only a subset of which actually commit to and signal the final decision.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified
How does fMRI deconvolution overcome the low temporal resolution of blood oxygen level-dependent (BOLD) signals?
It mathematically reverses the hemodynamic blurring and delay, recovering an approximation of the fast, underlying neural activity profile.
Source: Accumulation System: Distributed Neural Substrates of Perceptual Decision Making Revealed by fMRI Deconvolution · DOI · quote verified