Does zapping visual cortex at the right moment slow decisions?
Briefly disrupting the lateral occipital cortex 400 ms after a hard-to-see image appeared slowed people's face-or-car decisions, just as a prior brain-network model predicted.
Source
Using Transcranial Magnetic Stimulation to Test a Network Model of Perceptual Decision Making in the Human Brain
Study at a glance
- Design
- Human experiment — Within-subject chronometric paired-pulse TMS to left and right lateral occipital cortex at five timings during a difficult face/car discrimination, with a 500 ms timing as the control.
- N
- N=13 · 13 healthy adults completed both TMS sessions (15 recruited, 2 dropped out).
- Population
- Healthy young adults screened for psychiatric and neurological history, at Duke University
- Outcome
- Reaction time and accuracy for face versus car decisions at each TMS timing
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Thirteen healthy adults judged whether noisy, briefly flashed images showed a face or a car, with image clarity adjusted for each person so they were right about 79% of the time. Across two sessions (one for each hemisphere), pairs of TMS pulses 50 ms apart were delivered over lateral occipital cortex at one of five times relative to the image (200 ms before, or 200, 400, 450 or 500 ms after), plus trials without TMS. Stimulation at 500 ms, after the model says decision processing is finished, served as the control.
What they found
As predicted, TMS starting at 400 ms slowed reaction times compared with the 500 ms control (Cohen's d = 0.84), with no difference between hemispheres or between faces and cars. Unexpectedly, TMS 200 ms before the image sped responses (d = 1.33), and TMS at 200 ms reduced accuracy for cars only. The 500 ms control did not differ from no-TMS trials, supporting its use as a baseline.
The limits
What it doesn't show
With only 13 participants the findings need replication. Targets were based on group fMRI coordinates rather than each person's own activation, and there was no sham or control brain site, so the speed-up before the image may reflect a non-specific effect of the click and tap of TMS (intersensory facilitation) rather than real enhancement. The study tests one prediction of the model; it does not show that lateral occipital cortex is itself the accumulator of evidence rather than feeding another region.
Key terms
- Transcranial magnetic stimulation (TMS)
- A non-invasive method that uses a magnetic pulse to briefly disrupt or excite activity in a targeted patch of cortex.
- Chronometric TMS
- Delivering TMS at different delays after a stimulus to find when a region is needed for a task.
- Lateral occipital cortex (LOC)
- Visual cortex on the side of the occipital lobe involved in recognising objects.
- Perceptual decision-making
- The process of accumulating sensory evidence to choose between interpretations of a stimulus.
- Drift diffusion model
- A model in which evidence accumulates noisily toward a decision boundary; a lower drift rate means slower decisions.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
At which TMS timing did reaction times slow, as predicted?
Common questions
Why use TMS rather than more EEG or fMRI?
Imaging only shows that activity accompanies a decision; disrupting a region with TMS at a specific moment and seeing behaviour change is evidence that the region is causally needed then.
Why is stimulation at 500 ms a good control?
It hits the same tissue and feels identical to the other conditions, but occurs after the modelled decision stage, so any effect at 400 ms can be attributed to timing rather than the sensation of TMS.
Why were images made hard to see?
The model predicts that harder discriminations delay the decision-related brain signal to around 400 ms, so difficulty was tuned to place the process in the targeted time window.
More on Perceptual decision