Katherine E. Macey

A method for removal of global effects from fMRI time series.

We present a technique for removing global effects from functional magnetic resonance imaging (fMRI) images, using a voxel-level linear model of the global signal (LMGS). The procedure does not assume low-frequency global effects and is based on the assumption that the global signal (the time course of the average intensity per volume) is replicated in the same pattern throughout the brain, although not necessarily at the same magnitude. A second assumption is that all effects that match the global signal are of no interest and can be removed. The method involves modeling the time course of each voxel to the global signal and removing any such global component from the voxels time course. A challenge that elicits a large change in the global blood oxygenation level-dependent (BOLD) signal, inspired hypercapnia (5% CO(2)/95% O(2)), was administered to 14 subjects during a 144-s, 24-scan fMRI procedure; baseline series were also collected. The method was applied to these data and compared to intensity normalization and low-frequency spline detrending. A large global BOLD signal increase emerged to the hypercapnic challenge. Intensity normalization failed to remove global components due to regional variability. Both LMGS and spline detrending effectively removed low-frequency components, but unlike spline detrending (which is designed to remove only low frequency trends), the LMGS removed higher-frequency global fluctuations throughout the challenge and baseline series. LMGS removes all effects correlated with the global signal, and may be especially useful for fMRI data that include large global effects and for generating detrended images to use with subsequent volume-of-interest (VOI) analyses.

Functional magnetic resonance imaging responses to expiratory loading in obstructive sleep apnea.

Obstructive sleep apnea (OSA) is characterized by diminished upper airway muscle phasic and tonic activation during sleep, but enhanced activity during waking. We evaluated neural mechanisms underlying these patterns with functional magnetic resonance imaging procedures during baseline and expiratory loading conditions in nine medication-free OSA and 16 control subjects. Both groups developed similar expiratory loading pressures, but appropriate autonomic responses did not emerge in OSA cases. Reduced neural signals emerged in OSA cases within the frontal cortex, anterior cingulate, cerebellar dentate nucleus, dorsal pons, anterior insula and lentiform nuclei. Signal increases in OSA over control subjects developed in the dorsal midbrain, hippocampus, quadrangular cerebellar lobule, ventral midbrain and ventral pons. Fastigial nuclei and the amygdala showed substantially increased variability in OSA subjects. No group differences were found in the thalamus. OSA patients show aberrant responses in multiple brain areas and inappropriate cardiovascular responses to expiratory loading, perhaps as a consequence of previously-demonstrated limbic, cerebellar and motor area gray matter loss.

Inspiratory loading elicits aberrant fMRI signal changes in obstructive sleep apnea.

We hypothesized that neural processes mediating deficient sensory and autonomic regulatory mechanisms in obstructive sleep apnea (OSA) would be revealed by responses to inspiratory loading in brain regions regulating sensory and motor control. Functional magnetic resonance imaging (fMRI) signals and physiologic changes were assessed during baseline and inspiratory loading in 7 OSA patients and 11 controls, all male and medication-free. Heart rate increases to inspiratory loading began earlier and load pressures were achieved later in OSA patients. Comparable fMRI changes emerged in multiple brain regions in both groups, including limbic, cerebellar, midbrain, and primary motor cortex. However, in OSA subjects, altered signals appeared in primary sensory thalamus and sensory cortex, supplementary motor cortex, cerebellar cortex and deep nuclei, cingulate, medial temporal, and insular cortices, right hippocampus, and midbrain. Signal delays occurred in basal ganglia. We conclude that areas mediating sensory and autonomic processes, and motor timing, are affected in OSA; many of these areas overlap regions of previously demonstrated gray matter loss.

FMRI Responses to Hyperoxia in Congenital Central Hypoventilation Syndrome

Congenital Central Hypoventilation Syndrome (CCHS) patients show partial retention of peripheral chemoreception despite impaired ventilatory responses to CO2 and hypoxia. The condition allows examination of central responses to hyperoxia, which minimizes afferent traffic from peripheral chemoreceptors. We used functional magnetic resonance imaging to assess blood oxygen level–dependent signals over the brain during a baseline and subsequent 2-min hyperoxia (100% O2) period in 14 CCHS and 15 control subjects. After partitioning gray matter and correcting for global effects, the images were analyzed using volume-of-interest time trends followed by repeated-measures ANOVA and conventional cluster analyses. Respiratory rates initially (first 20 s) fell in CCHS, but rose in control subjects; CCHS heart rate increased in the first minute, and then decreased in the second minute, as in controls, but with muted rise and extent of decline. Multiple sites within the cerebellum, midbrain, and pons responded similarly to the challenge in both groups. Response patterns differed early in the right amygdala, paralleling initial respiratory pattern deficits, and late in the right insula, concomitant with cardiac rate differences. Signals also differed between groups in the medial and anterior cingulate, hippocampus, head of caudate, and lentiform nuclei, as well as pontine and midbrain structures and regions within the superior temporal and inferior frontal cortical gyri. The findings emphasize that structures that can alter respiratory timing, such as the amygdala, and modulate sympathetic outflow, such as the right insula, are deficient in CCHS. Medullary and pontine areas targeted by PHOX2B expression are also affected.

Aberrant neural responses to cold pressor challenges in congenital central hypoventilation syndrome.

Patients with congenital central hypoventilation syndrome (CCHS), a condition characterized by impaired ventilatory responses to chemoreceptor stimulation, do not show the normal increase in respiratory rate and respiratory-related heart rate variation to cold forehead stimulation, a challenge that bypasses central chemoreceptors. We hypothesized that a forehead cold pressor challenge would reveal abnormal neural response patterns, as assessed by functional magnetic resonance imaging, in brain regions that are responsible for the integration of cold afferent stimulation with respiratory and cardiovascular output in patients with CCHS. Primary sensory thalamic and cortical areas for the forehead showed diminished responses in 13 patients with CCHS (ventilator dependent during sleep but not waking, no Hirschsprungs disease) compared with 14 control subjects, despite initial signal changes in the cortex being similar in both groups. Cerebellar cortex and deep nuclei; basal ganglia; and middle to posterior cingulate, insular, frontal, and temporal cortices showed reduced signal rises in patients with CCHS. Areas within the frontal and anterior cingulate cortices exhibited marked signal declines in control subjects but little change in patients with CCHS. No response occurred in either group in the dorsal medulla, but medial and ventral medullary areas showed enhanced signals in patients with CCHS. The cold pressor stimulation did not recruit dorsal medullary sites that would be affected by PHOX2B (a mutation of which is associated with the syndrome) expression in either group but demonstrated deficient cerebellar and medial medullary influences that, by action on rostral sites, may underlie the loss of respiratory responses.

Global BOLD MRI changes to ventilatory challenges in congenital central hypoventilation syndrome

We evaluated global blood oxygen level dependent (BOLD) signal changes in gray and white matter in 14 congenital central hypoventilation syndrome (CCHS) and 14 control subjects. One baseline image series with room air and three series with 30 s room air followed by 120 s hypercapnia (5% CO2/95% O2), hypoxia (15% O2/85% N2) or hyperoxia (100% O2) were collected. Hypercapnia and hyperoxia raised, and hypoxia lowered gray and white matter global signal in both groups, with smaller changes in white matter. Signal changes in CCHS cases were lower than control subjects for hypercapnia in gray and white matter, slightly more-enhanced in hypoxia, and, except for initial transient responses, were nearly comparable during hyperoxia. Initial signal rate or pattern changes emerged in all three challenges in gray or white matter in control, but not CCHS cases. Neural or vascular mechanisms mediate perfusion differently in CCHS; the aberrant initial transient responses may reflect deficiencies in rapidly-varying physiologic interactions in the syndrome.

Regional brain response patterns to Cheyne-Stokes breathing

Cheyne-Stokes breathing (CSB) results from impaired integration of sensory information with respiratory motor output; however, regions mediating the disturbed control are unknown. We examined functional magnetic resonance imaging signals during CSB within sleep to determine affected areas. Two male patients with severe obstructive sleep apnea were scanned while asleep over multiple sessions during which they exhibited CSB. Significant signal increases coincident with apneic periods emerged bilaterally in the cerebellar cortex, hippocampus, mediodorsal thalamus, frontal cortex and precentral gyrus. Signals declined bilaterally in the anterior cingulate cortex and postcentral gyrus. The reduced activation in primary sensory cortex and increased signals prior to breathing onset in the motor cortex are consistent with loss of sensory stimulation by airflow, and with anticipatory action of the motor cortex prior to initiation of breathing. Hippocampal and anterior cingulate cortex participation likely reflect previously-demonstrated roles for initiating inspiratory efforts and resolving sensory information and motor action, respectively.

Wavelet median denoising of ultrasound images

Ultrasound images are contaminated with both additive and multiplicative noise, which is modeled by Gaussian and speckle noise respectively. Distinguishing small features such as fallopian tubes in the female genital tract in the noisy environment is problematic. A new method for noise reduction, Wavelet Median Denoising, is presented. Wavelet Median Denoising consists of performing a standard noise reduction technique, median filtering, in the wavelet domain. The new method is tested on 126 images, comprised of 9 original images each with 14 levels of Gaussian or speckle noise. Results for both separable and non-separable wavelets are evaluated, relative to soft-thresholding in the wavelet domain, using the signal-to-noise ratio and subjective assessment. The performance of Wavelet Median Denoising is comparable to that of soft-thresholding. Both methods are more successful in removing Gaussian noise than speckle noise. Wavelet Median Denoising outperforms soft-thresholding for a larger number of cases of speckle noise reduction than of Gaussian noise reduction. Noise reduction is more successful using non-separable wavelets than separable wavelets. When both methods are applied to ultrasound images obtained from a phantom of the female genital tract a small improvement is seen; however, a substantial improvement is required prior to clinical use.

Calculating rhythmicity of infant breathing using wavelets

Breathing signals are one set of physiological data that may provide information regarding the mechanisms that cause SIDS. Isolated breathing pauses have been implicated in fatal events. Other features of interest include slow amplitude modulation of the breathing signal, a phenomenon whose origin is unclear, and periodic breathing. The latter describes a repetitive series of apnea, and may be considered an extreme manifestation of amplitude modulation with successive cessations of breathing. Rhythmicity is defined to assess the impact of amplitude modulation on breathing signals and describes the extent to which frequency components remain constant for the duration of the signal. The wavelet transform was used to identify sections of constant frequency components within signals. Rhythmicity can be evaluated for all the frequency components in a signal, for individual frequencies. The rhythmicity of eight breathing epochs from sleeping infants at high and low risk for SIDS was calculated. Initial results show breathing from infants at high risk for SIDS exhibits greater rhythmicity of modulating frequencies than breathing from low risk infants.

Detecting variable responses in time-series using repeated measures ANOVA: Application to physiologic challenges.

We present an approach to analyzing physiologic timetrends recorded during a stimulus by comparing means at each time point using repeated measures analysis of variance (RMANOVA). The approach allows temporal patterns to be examined without an a priori model of expected timing or pattern of response. The approach was originally applied to signals recorded from functional magnetic resonance imaging (fMRI) volumes-of-interest (VOI) during a physiologic challenge, but we have used the same technique to analyze continuous recordings of other physiological signals such as heart rate, breathing rate, and pulse oximetry. For fMRI, the method serves as a complement to whole-brain voxel-based analyses, and is useful for detecting complex responses within pre-determined brain regions, or as a post-hoc analysis of regions of interest identified by whole-brain assessments. We illustrate an implementation of the technique in the statistical software packages R and SAS. VOI timetrends are extracted from conventionally preprocessed fMRI images. A timetrend of average signal intensity across the VOI during the scanning period is calculated for each subject. The values are scaled relative to baseline periods, and time points are binned. In SAS, the procedure PROC MIXED implements the RMANOVA in a single step. In R, we present one option for implementing RMANOVA with the mixed model function “lme”. Model diagnostics, and predicted means and differences are best performed with additional libraries and commands in R; we present one example. The ensuing results allow determination of significant overall effects, and time-point specific within- and between-group responses relative to baseline. We illustrate the technique using fMRI data from two groups of subjects who underwent a respiratory challenge. RMANOVA allows insight into the timing of responses and response differences between groups, and so is suited to physiologic testing paradigms eliciting complex response patterns.