| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 80 No. 3 September 1998, pp. 1522-1532
Copyright ©1998 by the American Physiological Society
Department of Neurology, Laboratory of Neuro Imaging, Division of Brain Mapping, University of California at Los Angeles School of Medicine, Los Angeles, California 90095-1769
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Cannestra, Andrew F., Nader Pouratian, Marc H. Shomer, and Arthur W. Toga. Refractory periods observed by intrinsic signal and fluorescent dye imaging. J. Neurophysiol. 80: 1522-1532, 1998. All perfusion-based imaging modalities depend on the relationship between neuronal and vascular activity. However, the relationship between stimulus and response was never fully characterized. With the use of optical imaging (intrinsic signals and intravascular fluorescent dyes) during repetitive stimulation paradigms, we observed reduced responses with temporally close stimuli. Cortical evoked potentials, however, did not produce the same reduced responsiveness. We therefore termed these intervals of reduced responsiveness "refractory periods." During these refractory periods an ability to respond was retained, but at a near 60% reduction in the initial magnitude. Although increasing the initial stimulus duration lengthened the observed refractory periods, significantly novel or temporally spaced stimuli overcame them. We observed this phenomenon in both rodent and human subjects in somatosensory and auditory cortices. These results have significant implications for understanding the capacities, mechanisms, and distributions of neurovascular coupling and thereby possess relevance to all perfusion-dependent functional imaging techniques.
The concept that changes in the level of neuronal activity can cause variations in the caliber of the cerebral vasculature is over 100 years old (Roy and Sherington 1890 We utilized repetitive stimulus paradigms to investigate response capacities over human and rodent cortex. An initial stimulus was presented, followed by a quiescent period of no stimulation (interstimulus interval) and a subsequent secondary stimulus. The duration of the secondary stimulus was equal to the initial stimulus. In rodents, whisker C1 was stimulated via an electromechanical motorized nudger, angle of deflection 30°, velocity 1.2 m/s anterior to posterior (or inferior to superior). Auditory stimulations of 10 kHz were provided to rodents utilizing a computer-controlled tone generator. Human transcutaneous electrical stimuli consisted of 17.5-mA pulsed current applied to the median nerve with the use of a bipolar electrode and Grass Instruments stimulator (S-12, Quincy, MA). Interstimulus intervals (0.1-15 s) and stimulus durations (0.5-15 s) were varied as well as stimulus frequency (5, 10, or 20 Hz). Two to 3 min separated each control/stimulation experiment set. Nonstimulation control trials were interleaved during all acquisitions (human and rodent). These control trials were performed immediately before stimulation experiments. During all animal experiments, single stimulation control studies were performed before repetitive paradigms were utilized. Single stimulation time courses were consistent with previous reports (Cannestra et al. 1996 Rodent OIS
Sixty-five adult male Sprague Dawley rats (200-450 g, Charles River Laboratories) were studied in accordance with Animal Research Committee institutional guidelines. Not all experimental trials were performed in all animals. Of the 65 animals utilized in this study, many were used to confirm the refractory period durations reported in Table 1. Subjects were anesthetized with halothane for femoral vein cannulation and subsequently maintained on a combination of halothane and urethan (initial dose 600 mg/kg) for the remainder of surgery. During imaging, urethan was administered (150-mg/kg dosage) to maintain anesthesia but to allow toe pinch withdrawal. Although the use of an anesthetic agent may affect these signals, we attempted to minimize this effect by frequent assessment and control of depth of anesthesia. Assessment was performed by monitoring toe pinch, corneal reflex, and respiration rate. Similar methods were used in previous publications (Blood et al. 1995
Rodent IVF dyes
After OIS imaging, 10 animals were given Texas red dextran polymer IVF dye (Molecular Probes; 50-70 mg/kg iv in physiological saline via syringe pump, 0.1 ml/min) to image blood volume changes. Details were provided in previous publications (Narayan et al. 1995 Evoked potentials
EPs were measured in rodents during repetitive C1 whisker stimulation utilizing paradigms identical to those used during OIS acquisitions. After optical imaging, a craniotomy was performed over somatosensory cortex. Recordings were made superficial to the cortical surface with the use of 100-µm tip diameter SNEX-200 bipolar electrodes (250-µm tip separation; Rhodes Medical, CA), placed stereotaxically corresponding to barrel C1. EPs were recorded from a Model 7D polygraph (60-Hz bandpass filter on, 3-Hz 1/2A low filter, 75-Hz 1/2A high filter; Grass Instruments, MA) and Nicolet Model 527 signal averager (Nicolet Instrument, Madison, WI). EPs were measured during 2- and 10-s stimulus paradigms. EPs during the first 16 2-s stimulus or first 32 10-s stimulus whisker deflections were measured, digitized, and averaged between six to eight trials per animal.
Human intraoperative optical imaging of intrinsic signals (iOIS)
We measured optical reflectance changes (utilizing paradigms as above) over sensorimotor cortex in three anesthetized patients undergoing surgical resection of frontoparietal tumors. Presurgical informed consent was obtained from all patients. Methodological detail was provided in previous publications (Toga et al. 1995 Optical image analysis
Subtraction and ratio image analysis [(stimulus-control)/(control)] was utilized to produce functional maps of activity (Narayan et al. 1994a
Evidence for reduced responsiveness
Figure 1 illustrates reduced responsiveness to temporally close stimuli. Repetitive stimulation paradigms were employed to characterize optical signals after an initial response (Fig. 1, top panel). Optical signals were recorded over rat somatosensory cortex in response to whisker (C1) deflection. Utilizing a 2-s stimulation paradigm (2-s initial and secondary stimulus) with a 3-s interstimulus interval, secondary responses were reduced 56 ± 19% (integral of curve; 76 ± 10% for intensity and 32 ± 12% for spatial involvement). During this 3-s interstimulus interval (between stimulations), the optical signal returned to prestimulus baseline levels. Increasing the interstimulus interval to 4 s resulted in secondary responses of equal magnitude (both intensity and spatial extent) to initial responses (Fig. 1, bottom panel). Single stimulation control experiments (Fig. 1, bottom panel, inset) provided temporal profiles identical to the primary response curves in Fig. 1, top and bottom panels. Single stimulation responses returned to and maintained a baseline after 6 s.
Characterization of diminished signals
A triple stimuli (2-s) paradigm was used to investigate further attenuation or recovery of diminished responses. Repetitive stimuli were introduced coincident with the observed recovery of OIS responses (4 s, as observed in Fig. 1). No changes in secondary or tertiary responses were observed, suggesting that the diminished signals may result from an inability of initial response to adapt to close temporally spaced stimuli.
Spatial analysis
ROI, intensity, COM, and principal axis calculations were utilized to examine spatial patterns during reduced responsiveness. Overall, cortical optical responses decreased to 34 ± 15% from initial spatial involvement (100%) and 63 ± 11% from initial intensity. The spatial reductions during the refractory period were nearly concentric, resulting in more focal secondary responses (Fig. 4). This result was confirmed by COM calculations that varied only within SE (120 µm) between primary and secondary responses. COM always localized within 1 mm of the stereotaxic position of C1. Principal axes were always aligned orthogonal with the major axis oriented along the posterior-inferior feeding branch of the MCA (for barrel cortex). A similar spatial response pattern was observed during IVF dye experiments (29% involvement of primary response). Auditory cortex demonstrated a similar spatial pattern, with reduction to 21%. During the frequency modulation experiments (Fig. 7), the initial peak and secondary peak (associated with frequency change) demonstrated similar patterns with spatial involvements and COM equal with SE (19% and 150 µm, respectively).
The observations of this study found reduced functional responses (both OIS and IVF dye) in the presence of equivalent neuronal activations (cortical EPs). Diminished functional responses were observed over rodent somatosensory and auditory cortices as well as sensorimotor cortex in humans. Stimulus duration, frequency, orientation, and temporal proximity dramatically affected the degree of reduced responsiveness. These results may have significant implications for perfusion-based imaging studies.
Evidence for refractory periods
Because this reduced responsiveness resembles the refractory nature of other physiological systems, we chose the term refractory period to describe this response pattern. Figures 1-3 demonstrate a return to prestimulus baseline response between stimulations. During single stimulus control experiments (Fig. 1, bottom panel), baseline levels were maintained for the entire poststimulus trial duration. Refractory periods may therefore not represent response overlap (positive or negative), but rather may demonstrate dependence on recent activity. Cortical EP measurements (Figs. 1 and 4), however, were equivalent during diminished responses. Refractory periods may therefore indicate processes indirectly coupled to neuronal firing.
Implications
Significant periods of reduced responsiveness (4 s) were observed (Table 1) for short activations (0.5 s). Refractory periods increased when stimulus duration was maintained during and beyond the OIS signal peak (i.e., >2.45 s). Modulating stimulation frequency did not change refractory period durations, suggesting that peripheral receptor tuning curves and intrinsic firing rates did not contribute to the observed stimulus dependence.
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
). Changes in functional perfusion were exploited to map cortical activation (Phelps and Mazziotta 1985), forming the basis for most modern functional neuroimaging techniques, including functional magnetic resonance imaging (fMRI), positron emission tomography (PET), single photon emission computed tomography (SPECT), optical imaging of intrinsic signals (OIS), near infrared spectroscopy (NIRS), and transcranial Doppler ultrasonography (TCD) (Grinvald et al. 1986; Hurtig et al. 1994; Kwong et al. 1992; Lassen and Holm 1992; Sitzer et al. 1994; Villringer et al. 1993). These methods assume tight coupling between perfusion and neuronal activity. The full characterization of neurovascular coupling is therefore critical for our understanding and quantification of cortical activity with the use of perfusion-based methods. OIS currently offers the spatial (µm) and temporal (ms) resolution necessary to observe rapid and transient signal changes by measuring changes in blood volume, cellular swelling, hemoglobin concentrations, and cytochrome activity (Cohen 1973; Holthoff and Witte 1996; Malonek and Grinvald 1996; Narayan et al. 1995). Optical signals were first demonstrated in vitro by Hill and Keynes (1949) and were used for functional imaging in rodents (Blood et al. 1995), primates (Frostig et al. 1990), and humans (Haglund et al. 1992; Toga et al. 1995). Optical signals colocalize with electrically active cortex (Narayan et al. 1994a,b) and cytochrome oxidase stained barrels in rodent (Blood et al. 1995). The etiology of the intrinsic signals includes reflectance changes from several optically active processes (vascular and metabolic in origin) (Cohen 1973), depending on the wavelength measured (Malonek and Grinvald 1996) rather than indicating neuronal firing directly. Narayan et al. (1995) provided evidence that intrinsic signals are closely related to increases in cerebral blood volume. Optical responses in rodent and human somatosensory cortex were characterized describing nonlinear, overshoot, and delay phenomena (Cannestra et al. 1996) in good agreement with other measures of vascular responsivity (Ngai et al. 1988, 1995; Sitzer et al. 1994).
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
).
; Cannestra et al. 1996). Briefly, animals were stabilized in a stereotaxic frame, heads shaved, and skull cleaned. A scraping instrument (Biomedical Research Instruments, Rockville, MD) was used to uniformly thin bone (near 250 µm) over right primary somatosensory or auditory regions. Silicon oil was applied to increase bone translucency. Imaging was performed (50- to 200-ms exposure, 192 × 144 pixel array) with a slow scan CCD camera (model TE/CCD-576EFT, Princeton Instruments, Trenton, NJ) mounted atop a microscope. Voltage-stabilized white light was used to epi-illuminate cortex. Images were acquired through transmission filters at 550 (540-560) nm, 650 (640-660) nm, and 850 (840-860) nm (Corion, Holliston, MA).
View this table:
TABLE 1.
Effects of stimulus duration on interval of reduced responsiveness
). The large molecular weight (MW, 70,000) restricts Texas red distribution to the vasculature. During image acquisition (as described previously) the cortex was epi-illuminated with 590-nm light (580-600 nm filter, Corion). Cortical reflectance emission was filtered at 650 nm. IVF dye images were acquired 5 min after onset of dye infusion to allow for equilibration. Stimulation paradigms were identical to those used during OIS acquisitions.
). The patient's head was fixed, via a Mayfield apparatus, to the operating table. After site localization (via a frameless stereotaxic system), the craniotomy and reflection of the dura was performed.
). Control images were randomly selected from the nonstimulated interleaved control trials. Control and stimulated images were averaged across three to five trials to increase signal to noise (SNR) (Janesick et al. 1987).
; Geladi et al. 1989; Yap et al. 1994). Briefly, given a set of images [Xt=0
N] (image size 192 × 144) from time points 0 to N, we construct a single matrix [XS] (size 27,648 × N, 192 × 144 = 27,648, 0 to N 
1 = N time points) in which the columns represent a pixel intensity for each image and the rows represent the pixel intensity time course. All pixels are utilized [no regions of interests (ROIs) are specified]. Matrix [XS]T is transposed and the covariance matrix [C] (11 × 11) is determined as [C] = [XS] [XS]T. The eigenvalues (
i, where i = 1 N) and eigenvectors (V
i) are determined by the Jacobi transformations technique. Because PCA is not biased by the determination of spatial constraints (such as ROIs), the principal component is an appropriate representation of temporal functions and was used to measure of OIS time courses.
View larger version (32K):
[in a new window]
FIG. 1.
Existence of refractory behavior. Principal component analyses (PCAs) are presented representing temporal response (Cannestra et al. 1996 ). The first eigen vector represents 93-99% of the observed time courses in the optical images and is region of interest (ROI) independent (i.e., utilizes the entire image). This unitless quantity is well suited for averaging across subjects and species to obtain an average multisubject response profile. Multiple (9) trials from 3 animals were averaged and displayed for each graph. Optical responses (850 nm) were recorded after C1 vibrissal stimulation (10-Hz deflection) in a repetitive stimulation paradigm (solid line). An initial stimulus was presented to the animal, followed by a quiescent period of no stimulation (interstimulus interval) and a subsequent secondary stimulus. The experiments of Fig. 1 were first performed at a temporal resolution of 750 ms. After the initial characterization of the refractory period utilizing a 2-s stimulation paradigm (based on 6 subjects), the experimental paradigm was repeated at a higher resolution (200 ms) to examine for transient events contributing to the refractory period behavior. High-resolution experiments (based on 3 animals) provided identical results to the previous lower temporal resolution experiments. Horizontal bars indicate stimulation durations. Interleaved nonstimulated control trials are plotted (dashed line) right to left (time axis is seconds before experimental trial). Error bars indicate SE. Top: primary stimulus, 2.00 s; interstimulus interval, 3.00 s; secondary stimulus, 2.00 s. Commencement of the signal occurred by 1.00 s, peaked at 3.00 s, decayed until 5.25 s, rose to peak secondary response at 7.50 s, and decayed by 11.25 s. Peak secondary response is 56 ± 19% [integral of curve; 76 ± 10% (SE) for intensity and 32 ± 12% for spatial involvement] of primary response. Optical response returned to baseline during the interstimulus interval. Top, inset: EPs were recorded during the first 8 whisker deflections during primary (black) and secondary (shaded) stimulation. Recordings were made superficial to the cortical surface, stereotaxically corresponding to barrel C1. EPs were equivalent. Bottom: primary stimulus, 2.00 s; interstimulus interval, 4.00 s; secondary stimulus, 2.00 s. Commencement of the signal occurred by 1.00 s, peaked at 3.00 s, decayed until 6.00 s, rose to peak secondary response at 9.00 s, and decayed by 12.00 s. Peak secondary response is within SEs of primary response. Bottom, inset: optical imaging of intrinsic signals (OIS) response to a single 2-s stimulation. Response returned to and maintained baseline after 6 s. Poststimulus undershoots were not observed.
21) by the sum of squared eigenvalues (
2i, where i = 1 N), i.e., (21)/(
2i, where i = 1 N). The first eigenvalue (corresponding to the principal component) averaged 100 times and 30 times larger than the second eigenvalue for rodents and humans, respectively.
; Cannestra et al. 1996). The time point with the largest calculated magnitude for each peak was considered to be the maximal optical response. This "maximal" ROI was then superimposed upon the other images of the trial to calculate the respective magnitude at each time point.
Where gij is the gray scale value of the image pixel at the location (i,j). The principal axes of the image were defined as the eigenvectors of the 2 × 2 inertia matrix M where
Similar calculations are commonly used in functional imaging and image coregistration (Toga and Banerjee 1993).
View larger version (21K):
[in a new window]
FIG. 2.
Refractory periods were observed at multiple wavelengths of optical imaging (540, 650, and 850 nm). The principal components are displayed as an indicator of temporal response. Optical responses were recorded utilizing the paradigms in Fig. 1, top, through a single wavelength transmission filter (540, 650, or 850 nm). OIS at all 3 wavelengths produced similar temporal profiles. The optical signals responded within 1 s of stimulation, peaked at 3 s, returned to baseline ~7 s, and produced a reduced secondary peak at 8 s poststimulation. Secondary peak responses were reduced 26, 24, and 22% (integral of curve; 55, 66, and 70% for intensity and 40, 18, and 26% for spatial involvement) for 540, 650, and 850, respectively. Time courses are single trials produced sequentially from a single subject (3 trials averaged for each graph). All wavelengths demonstrated an initial response peak ~2.5 s and a secondary reduced response peak ~8 s. Horizontal bars, stimulus durations; nonstimulated control trials are plotted in the inset (time axis is seconds before experimental trial). Trial-to-trial variabilities were ~11, 15, and 13% for 540, 650 and 850 nm, respectively.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
; Narayan et al. 1995).
View larger version (90K):
[in a new window]
FIG. 3.
Refractory periods also were observed utilizing intravascular fluorescent (IVF) dyes. After OIS imaging, Texas red dextran polymer fluorescent dye was infused for blood volume fluorescence imaging. Fluorescence at 650 nm was measured utilizing 590-nm source (excitation) light. The same paradigm as in Figs. 1 and 2 was used. Fluorescence commenced by 1 s, increased to peak ~3 s, reduced to near baseline by 6 s, and rose to a 2nd peak ~8 s. Secondary fluorescence is diminished 26% (integral of curve; 43% for intensity and 29% for spatial involvement) from primary response (3- vs. 7.4-s images), correlating temporally (see Fig. 1) and spatially with OIS images. OIS activation of 850 nm at 3.0s is displayed at bottom right. Vessels observed in the images are branches of the middle cerebral artery. Raw fluorescence image is at middle right. Images are original data, pseudocolored and displayed with 3-pixel Gaussian blur. Images and principle components are from a single subject (six trials averaged). The principal components (solid line, stimulation trial; dashed line, control trial) are displayed as a time-course measure. The nonstimulated control is plotted right to left (time axis is seconds before experimental trial). A, anterior; L, lateral. Top color bar, fluorescence increase ×10 3; scale bar, 1 mm; horizontal bars, stimulus durations. Trial-to-trial variability was ~8%. Bottom color bar, reflectance decrease ×104.
; Ngai et al 1988 1995; Sitzer et al. 1994; Yamashita et al. 1996) demonstrated an initial peak (Fig. 4A) reducing to a steady-state response (maintenance level, as in Fig. 4B) during sustained stimulation. Utilizing repetitive paradigms combined with sustained stimulation (10-15 s), we observed an absence of the initial peak. Introduction of a second stimulus initiated a response that only attained the maintenance level observed during primary stimulation, foregoing the initial peak phase (Fig. 4C). The observed reduced responsiveness therefore may only apply to the initial response peak. Consistent with previous reports (Cannestra et al. 1996), single stimulation control experiments provided temporal evolutions identical to the primary response curve in Fig. 4.
View larger version (51K):
[in a new window]
FIG. 4.
Observed loss of the initial response peak. Optical response of 850 nm was measured (as in Fig. 1) from a primary stimulus duration of 10.00-s, interstimulus interval of 3.50-s, and secondary stimulus 10.00-s paradigm. Principal components (12 trials from 6 animals) are plotted as a time-course measure. Commencement of the signal (solid line) occurred by 1.00 s, peaked at 3.75 s, decayed to maintenance level by 7.50 s, reduced to near baseline at 14.25 s, rose to maintenance level 2nd response ~15.75 s and decayed by 29.00 s. Error bars (SEs) during phases B and C overlap at respective time points indicating similarity of maintenance phases. Magnitude comparison of peak (initial response peak) with maintenance phase response is in agreement with the literature (Cannestra et al. 1996 ; Ngai et al. 1988, 1995; Sitzer et al. 1994; Yamashita et al. 1996). Interleaved nonstimulated control trials are plotted (dashed line) right to left (time axis is seconds before experimental trial). Inset: EPs were measured during the 1st 3.5 s of the 1st (blue) and 2nd (red) stimulations (synchronized to the beginning of whisker deflection at 0 ms; 6 trials averaged). Recordings were made superficial to the cortical surface, stereotaxically corresponding to barrel C1. Vessels observed in the images are branches of the middle cerebral artery. Images are original data, pseudocolored and displayed with 1-pixel Gaussian blur. Images and EPs are from a single subject. A, anterior; L, lateral. Color bar, reflectance decrease ×104; scale bar, 1 mm; all error bars are SEs; horizontal bars, stimulus durations.
View larger version (92K):
[in a new window]
FIG. 5.
Refractory periods also were observed over auditory cortex utilizing 10-kHz audible stimulation. Audible stimulation utilized the same paradigms as Figs. 1 and 2. Signal (850 nm) commenced by 1 s, increased to peak ~3 s, reduced to near baseline by 5.5 s, and rose to a 2nd peak ~7 s. Secondary response is diminished relative to primary response (3- vs. 6.75-s images). Peak response was reduced to 26% (integral of curve; 39% for intensity and 21% for spatial involvement). The nonstimulated control trial is plotted (dashed line) right to left (time axis is seconds before experimental trial). A schematic indicating field of view and orientation is provided (bottom right). Images are original data, pseudocolored and displayed with 1-pixel Gaussian blur. Images are from a single subject (six trials averaged). Color bar, reflectance decrease ×10 4; scale bar, 1 mm; trial-to-trial variability was ~11%.
View larger version (20K):
[in a new window]
FIG. 6.
Evidence for temporal dynamic behavior and vascular refractory periods in humans. Optical responses (610 nm) were recorded during a primary stimulus (transcutaneous median nerve stimulation, 17.5 mA) through a quiescent interstimulus interval and a subsequent secondary stimulus (same stimulus parameters). The principal components are displayed as a measure of temporal response. Time courses are individual stimulation trials (solid line) from a single subject. Interleaved nonstimulated control trials are plotted (dashed line) right to left (time axis is seconds before experimental trial). A: stimulus durations (horizontal bars) are 2 s. Interstimulus interval is 3 s. Optical response began at a prestimulus baseline, rose to initial peak at 2.0 s, reduced to near baseline at 5.0 s, rose to a 2nd peak at 7.0 s, and reduced to baseline by 13.0 s. Secondary response was ~80% magnitude of the first. B: stimulus durations are 10 s. Interstimulus interval is 4 s. Optical response began at a prestimulus baseline, rose to initial peak at 2.2 s (Ba), reduced to a maintenance phase (Bb) from 4.4 to 9.9 s, rose to a 2nd peak (Bc), and maintained this level from 17.6 to 22.0 s. Secondary response only attained the maintenance level observed during primary response (Bc). The observed initial peak is nearly 3 times the observed maintenance level.
View larger version (30K):
[in a new window]
FIG. 7.
Changing stimulus frequency induces an initial response peak during sustained stimulation. During a 25-s stimulation in rodents (10 trials in 5 animals), whisker deflection frequency was either doubled (A: 10-20 Hz) or halved (B: 10-5 Hz) at 13.5 s of stimulation. The principal components (3 animals) are displayed as a time-course measure (850-nm OIS). In both stimulation paradigms (solid lines), the change in stimulus frequency induced a 2nd response peak (Ab and Bb) of equal magnitude to the initial response peak (Aa and Ba), respectively. Reappearance of the initial response peak during increased (20 Hz; Aa) stimulation indicates the relative nature of refractory periods. Additionally, the peak observed during decreased frequency stimulation (5 Hz; Bb) suggests mechanisms responsible for qualitative changes (increases or decreases) in stimuli may overcome refractory periods. Interleaved nonstimulated control trials are plotted (dashed line) right to left (time axis is seconds before experimental trial). Error bars are SEs; horizontal bars are stimulus durations.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Grinvald et al. 1986; Narayan et al. 1994b), their appearance is variable (Narayan et al. 1994b) and may depend on wavelength (near 600 nm) (Frostig et al. 1990; Malonek and Grinvald; 1996). Poststimulus undershoots were not reliably observed at any wavelength utilized in this study (Fig. 1, inset, bottom panel).
), the etiology of these signals includes reflectance changes from several optically active processes (Cohen 1973; Holthoff and Witte 1996) considered wavelength dependent (Frostig et al. 1990; Malonek and Grinvald 1996). Scattering effects from increased intravascular volume, vascular morphologic changes, and cellular swelling are known to contribute to many regions of the visual spectrum (including 540 and 650 nm) and are assumed emphasized at longer wavelengths (i.e., 850 nm). Wavelengths near 650 nm reflect activity-dependent oxygen delivery from capillaries, whereas imaging near 540 nm indicates activity-dependent blood volume increases. Although the optical intrinsic signal is not an exclusive indicator of vascular response, observation of refractory periods at multiple wavelengths (540, 650, and 850 in rodents and 610 in humans) is consistent with optical contributions caused by regional cerebral blood flow changes (Frostig et al. 1990; Narayan et al. 1995). These time courses also are consistent with global signal light scattering components obtained from imaging spectroscopy (Malonek and Grinvald; 1996).
IVF dyes revealed reduced responsiveness to temporally close stimuli (Fig. 3). Both IVF dye signals and OIS signals were detected over cortical regions much larger than individual barrels. This finding was reported previously (Narayan et al. 1995) and may be caused by macrovascular contributions (venous and arterial), activity in lateral arbors of dendrites and axons, inhibitory activity (McCasland and Hibbard 1997), and propagating electrophysiological activity through adjacent layers and adjacent areas of layer IV (Armstrong-James et al. 1992). Although IVF dye and OIS maps colocalize over branches of the middle cerebral artery, the observed spatial differences may be related to the larger volume of tissue sampled at increased wavelengths (850 nm for OIS, 650 nm for IVF dyes). Previous studies colocalized multiwavelength optical signals (610 and 850 nm) and IVF dyes (Narayan et al. 1995). In this study, the temporal and spatial colocalization of IVF dye and OIS signals suggests refractory periods are related to blood volume changes (Narayan et al. 1995). Refractory periods may therefore represent dynamic behavior of neuronal and vascular coupling mechanisms. Additional support for a vascular etiology was revealed in OIS images where large signal contributions overlaid posterior branches of the middle cerebral artery (Figs. 3 and 4, A-C).
demonstrated a decrease in spatial involvement (both OIS and EPs) with increasing stimulation frequency. Because we were unable to detect electrophysiological differences between primary and secondary stimulation, refractory periods may be independent of the spatial/frequency relationship observed by Sheth et al. (1998).
demonstrated that considerable variation results from low-frequency oscillations of cerebral blood flow, which could further influence our measurements. However, we did not detect these oscillations directly.
). Although the human results presented here are limited because of intraoperative constraints, the temporal patterns and reduced responsiveness were consistent with rodent studies (Figs. 1 and 6A, 4 and 6B, respectively). Additionally, we also observed refractory periods across cortical regions (sensory and auditory). Therefore refractory periods are probably not caused by the specific architecture of the barrel cortex but rather may represent mechanisms common to mammalian sensory cortices.
reported time constants from gray matter blood oxygen level dependent fMRI signals in good agreement with published OIS studies. fMRI (Bandettini et al. 1997; Kruger et al. 1997) studies also have shown dynamics similar to other perfusion-dependent techniques (Cannestra et al. 1996; Ngai et al. 1988, 1995; Sitzer et al. 1994; Yamashita et al. 1996). Accordingly, we postulate the vascular refractory period is observable with fMRI and that it is attenuated and susceptible to field strength and resolution (voxel size). Recently, single-trial fMRI experiments demonstrated a "departure from linearity" when examining individual trial response (Dale and Buckner 1997). These fMRI signal departures may be related to refractory periods. Currently, human multimodality neuroimaging (preoperative fMRI and subsequent intraoperative OIS) is in progress to elucidate the exact significance of these events for fMRI. These fMRI studies provided similar results (Toga et al. 1997) but with a smaller difference (8-15%) between the primary and secondary responses (paradigm equivalent to Fig. 1, top panel). The attenuation may result from the venous origin for the T2* fMRI signal and/or temporal/spatial differences.
; Simons 1985), these results may indicate that small changes in cortical activity can lead to disproportionately large perfusion responses; recent fMRI reports support this hypothesis. An initial fMRI response peak and subsequent maintenance phase are observed during diffuse visual stimulation; however, during reversing checkerboard stimulation, the initial fMRI response peak level is sustained for the entire stimulus duration (Kruger et al. 1997). Additionally, Bandettini et al. (1997) also reported response peaks corresponding to cessation of stimuli.
); however, changes in functional perfusion occur on the order of seconds (LeManna et al. 1987; Narayan et al. 1995; Ngai et al. 1988, 1995; Sitzer et al. 1994). Refractory periods may represent an overcompensation in response to initial stimulation or an inability to sustain perfusion levels caused by capacity limitations. Both explanations predict the initial peak observed during prolonged stimulation studies (Bandettini et al. 1997; Cannestra et al. 1996; Ngai et al. 1988; 1995). However, the ability to respond to changes in the stimulus during refractory periods (Fig. 7), supports the hypothesis that there is an overcompensation to initial (sufficiently novel or temporally spaced) stimulation. This model is consistent with a metabolic substance to signal vascular response (Moncada et al. 1991) and the time course of vascular smooth muscle relaxation (Ignarro et al. 1981). It also is consistent with oscillatory "vasomotion" signals (Mayhew et al. 1996), nonlinear responses to simultaneous stimulations (Toga et al. 1995), and mathematical models of neurovascular coupling (Buxton and Frank 1997).
). VPm neurons respond without attenuation up to 5 Hz and slightly attenuate up to 10 Hz. In contrast, POm neurons begin to attenuate their firing at 5 Hz and strongly attenuate at frequencies >10 Hz. In our study, frequency of whisker deflection (5, 10, or 15 Hz) did not affect the observed reduced responses. Therefore the VPm and POm attenuation frequencies are probably not responsible for intervals of reduced responsiveness. Similarly, we did not observe correlations between anesthesia dosage and reduced responses. Although urethan anesthesia was shown to delay components (~13 ms) of evoked response (Simons et al. 1992), we observed refractory periods both in rodents and in humans utilizing different anesthetic agents. Anesthesia, therefore, also is unlikely to account for the reduced responsiveness in our study.
; Cohen et al. 1997; Dale and Buckner 1997), interstimulus intervals should be considered to prevent response reductions because of previous activity.
| |
ACKNOWLEDGEMENTS |
|---|
We thank J. S. Burton and A. J. Blood for assistance.
This work is supported by National Institutes of Health Grants MH/NS-52083, Training Program Neuroimaging MH-19950 (A. F. Cannestra), and Medical Scientist Training Program GM-08042 (N. Pouratian and M. Shomer).
| |
FOOTNOTES |
|---|
Address for reprint requests: A. W. Toga, Dept. of Neurology, Laboratory of Neuro Imaging, Division of Brain Mapping, UCLA School of Medicine, 710 Westwood Plaza, Room 4238, Los Angeles, CA 90095-1769.
Received 16 December 1997; accepted in final form 18 May 1998.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  T. R. Husson, A. K. Mallik, J. X. Zhang, and N. P. Issa Functional Imaging of Primary Visual Cortex Using Flavoprotein Autofluorescence J. Neurosci., August 8, 2007; 27(32): 8665 - 8675. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Simons, J. Chiu, O. V. Favorov, B. L. Whitsel, and M. Tommerdahl Duration-Dependent Response of SI to Vibrotactile Stimulation in Squirrel Monkey J Neurophysiol, March 1, 2007; 97(3): 2121 - 2129. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Fukuda, U. M. Rajagopalan, R. Homma, M. Matsumoto, M. Nishizaki, and M. Tanifuji Localization of Activity-dependent Changes in Blood Volume to Submillimeter-scale Functional Domains in Cat Visual Cortex Cereb Cortex, June 1, 2005; 15(6): 823 - 833. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Caesar, L. Gold, and M. Lauritzen From the Cover: Context sensitivity of activity-dependent increases in cerebral blood flow PNAS, April 1, 2003; 100(7): 4239 - 4244. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Blood, N. Pouratian, and A. W. Toga Temporally Staggered Forelimb Stimulation Modulates Barrel Cortex Optical Intrinsic Signal Responses to Whisker Stimulation J Neurophysiol, July 1, 2002; 88(1): 422 - 437. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. F. Cannestra, N. Pouratian, S. Y. Bookheimer, N. A. Martin, D. P. Becker, and A. W. Toga Temporal Spatial Differences Observed by Functional MRI and Human Intraoperative Optical Imaging Cereb Cortex, August 1, 2001; 11(8): 773 - 782. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Ogawa, T.-M. Lee, R. Stepnoski, W. Chen, X.-H. Zhu, and K. Ugurbil An approach to probe some neural systems interaction by functional MRI at neural time scale down to milliseconds PNAS, September 26, 2000; 97(20): 11026 - 11031. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
