Article Figures & Data
Figures
- Figure 1.
Longitudinal investigation of beta rhythms (12–20 Hz) in the LFP of area V4 with and without V1 input by selective ablation of V1. A, During task periods without transient sensory input, such as active fixation (Fix.) before the onset of a stimulus (Stim.), the LFP in visual cortex of behaving subjects is often dominated by rhythmic fluctuations in the beta frequency range (12–20 Hz). B, To investigate the role of bottom-up input for the generation of these beta oscillations, we recorded from midlevel area V4 with chronically implanted arrays before and after a targeted aspiration lesion in V1. The lesion was placed to eliminate the V1 representation of the horizontal meridian (HM) between ∼2–7° of visual eccentricities (red) while leaving the lower vertical meridian (VM) representation, close to lunate sulcus, intact (gray). C, Example coronal section of V1 showing extent of lesioned (red line) and intact (gray line) tissue for Monkey B. Note the loss of gray matter in the targeted area. D, Stimulus locations throughout the paper are labeled lesion stimulus (LS) for stimuli inside the lesion-affected visual space around the HM and control stimulus (CS) for stimuli outside, close to the VM. E, Behavioral performance in a detection task before and after V1 lesion with dot stimuli around the lesion stimulus location. Arrows indicate time period of LFP recordings.
- Figure 2.
Beta oscillations during active fixation in area V4 are preserved after selective removal of V1. A, Example prestimulus beta oscillations of the unfiltered V4 LFP before (black, left) and after V1 lesion (red, right). B, Absolute power spectrum of the prestimulus period before (black, n = 5 sessions, each comprising 1000–1300 trials) and after (red, n = 13 sessions) from an example V4 electrode in Monkey F, averaged across trials and sessions. Gray and red shadings represent SEM across sessions before and after lesion, respectively. C, Distribution of power in the beta frequency range (12–20 Hz, green shading in B) during prestimulus period before and after V1 lesion. Each dot represents the beta-band power at a recording site averaged across sessions. Filled and open symbols represent recording sites with significant or nonsignificant changes in power (p < 0.05, independent samples t test), respectively. Gray line indicates identical prelesion and postlesion power. D–F, Example beta oscillations (D), wavelet spectrum (E), and beta power distribution (F) for Monkey B before (n = 7 sessions) and after (n = 5 sessions) V1 lesion.
- Figure 3.
Visual stimulation in lesion-affected visual space induces strong beta (12–20 Hz) oscillations in V4 instead of decreasing beta power. A, Example LFP responses to a moving grating stimulus (high contrast, varying spatial frequency, and drift direction) in the lesion-affected part of visual space from a recording site in Monkey F before the V1 lesion. Upper row, Single trial LFP trace. Gray line indicates stimulus onset. Lower row, Session-averaged (n = 5), baseline-normalized wavelet power spectra from an example electrode in Monkey F (data as in Fig. 2). White circle represents peak latency within the beta frequency band (dashed lines). B, Same as A for data from after the V1 lesion (n = 11 sessions). Note the unusual stimulus-induced beta oscillations visible in both single trials and average. C, Baseline-normalized power spectra from example recording site in A averaged across 200–500 ms period and sessions for lesion stimulus, before (black) and after (red) V1 lesion. Shadings surrounding curves represent SEM across sessions. D, Distribution of beta power peak latency before and after V1 lesion. E, Distribution of baseline-normalized power in beta band before and after V1 lesion for lesion stimulus. Each dot represents a power value from one recording site averaged across the beta band, 200–500 ms period, and sessions. F, Same as E for the gamma frequency band (30–150 Hz). E, F, Open and filled symbols represent recording sites with nonsignificant and significant changes in power on the individual electrode level (p < 0.05, independent samples t test), respectively.
- Figure 4.
Reversal of stimulus-induced beta oscillation dynamics is restricted to visual stimulation in lesion-affected space. A, B, Session-averaged, baseline-normalized wavelet power spectra from an example electrode in Monkey B around visual stimulation with drifting grating stimulus (high contrast, varying spatial frequency, and drift correction) at the control stimulus location before (A, n = 7 sessions) and after (B, n = 4 sessions) the V1 lesion. C, Baseline-normalized power spectra from example recording site in A, B averaged across 200–500 ms period and sessions for lesion stimulus, before (black) and after (red) V1 lesion. Shadings surrounding curves represent SEM across sessions. D, Distribution of beta power peak latency before and after V1 lesion E, Distribution of baseline-normalized power in beta band before and after V1 lesion for lesion stimulus. Each dot represents a power value from one recording site averaged across the beta band, 200–500 ms period, and sessions. F, Same as E for the gamma frequency band (30–150 Hz). E, F, Open and filled symbols represent recording sites with nonsignificant and significant changes in power on the individual electrode level (p < 0.05, independent samples t test), respectively.
- Figure 5.
V1-independent beta oscillations are sensitive to motion. A, Time course of baseline-normalized power in beta (12–20 Hz) band from example recording site and sessions in Monkey F for moving (green) and static (blue) grating stimuli (high contrast, varying drift direction/orientation, and spatial frequency for both) at the lesion stimulus location. Shading represents SEM. B, Distribution of beta power values (200–500 ms) for moving versus static stimuli for lesion (left) and control stimulus (right) from Monkey F (data as in Figs. 3 and 4). C, Same as B for Monkey B.
