Performance of visually guided tasks using simulated prosthetic vision and saliency-based cues.

OBJECTIVE: The objective of this paper is to evaluate the benefits provided by a saliency-based cueing algorithm to normally sighted volunteers performing mobility and search tasks using simulated prosthetic vision. APPROACH: Human subjects performed mobility and search tasks using simulated prosthetic vision. A saliency algorithm based on primate vision was used to detect regions of interest (ROI) in an image. Subjects were cued to look toward the directions of these ROI using visual cues superimposed on the simulated prosthetic vision. Mobility tasks required the subjects to navigate through a corridor, avoid obstacles and locate a target at the end of the course. Two search task experiments involved finding objects on a tabletop under different conditions. Subjects were required to perform tasks with and without any help from cues. RESULTS: Head movements, time to task completion and number of errors were all significantly reduced in search tasks when subjects used the cueing algorithm. For the mobility task, head movements and number of contacts with objects were significantly reduced when subjects used cues, whereas time was significantly reduced when no cues were used. The most significant benefit from cues appears to be in search tasks and when navigating unfamiliar environments. SIGNIFICANCE: The results from the study show that visually impaired people and retinal prosthesis implantees may benefit from computer vision algorithms that detect important objects in their environment, particularly when they are in a new environment.

Saliency-based image processing for retinal prostheses.

We present a computationally efficient model for detecting salient regions in an image frame. The model when implemented on a portable, wearable system can be used in conjunction with a retinal prosthesis, to identify important objects that a retinal prosthesis patient may not be able to see due to implant limitations. The model is based on an earlier saliency detection model but has a reduced number of parallel streams. Results of a comparison between the areas detected as salient by the algorithm and areas gazed at by human subjects in a set of images show a correspondence which is greater than what would be expected by chance. Initial results for a comparison of the execution speed of the two algorithm models for each frame on the TMS320 DM642 Texas Instruments Digital Signal Processor suggest that the proposed model is approximately ten times faster than the original saliency model.

The structural brain correlates of cognitive deficits in adults with Klinefelter's syndrome.

CONTEXT: Adults with Klinefelter's syndrome (KS) are known to present disturbances of language skills and delayed learning abilities. OBJECTIVES: The aim of this study was to assess brain morphometry in KS and to correlate eventual volumetric changes with performance on neuropsychological tests. PATIENTS: Patients included 18 KS adults and 20 age-matched controls. METHODS: All participants underwent prospectively double-spin-echo brain magnetic resonance imaging and neuropsychological testing of verbal and nonverbal domains. On the axial stack of magnetic resonance imaging slices, regional brain volumes were measured either by automated segmentation (full brain, total cerebrospinal fluid, and ventricular volume) or manual drawing with help of a neuroanatomy atlas (frontal, temporal, and parietal lobes, gray matter component of the lobes, cerebellar hemispheres, and hippocampal complexes). RESULTS: KS patients performed significantly lower than controls on language-related tasks exploring verbal processing speed and verbal executive function. They were diagnosed with significant enlargement of ventricular volume and bilateral reduction of cerebellar hemispheres. Furthermore, after separation of participants according to handedness and after correction of regional brain volumes for atrophy, a significant reduction of left temporal lobe volume was found in KS compared with controls. Ventricular volume was inversely correlated with cognitive function, whereas left temporal lobe volume was positively correlated with language-related tasks. CONCLUSION: This study hypothesizes that supernumerary X-chromosome and/or congenital hypogonadism provoke structural alterations in the subcortical pathways involved in language processing, thus providing a neurobiological substrate for cognitive deficits in KS.

Segmentation of progressive multifocal leukoencephalopathy lesions in fluid-attenuated inversion recovery magnetic resonance imaging.

BACKGROUND AND PURPOSE: The authors compared the reproducibility of a manual and a semiautomated technique for the quantitation of white-matter lesions in magnetic resonance imaging (MRI). METHODS: Volumes of white-matter lesions were determined using fluid-attenuated inversion recovery MRI in 23 AIDS patients with progressive multifocal leukoencephalopathy. Manual outlining was compared to an automated method based on region growing and adaptive thresholding. RESULTS: Lesion volumes from the 2 methods correlated well (61 lesions, r = 0.99, P < 10(-4)), although the volumes differed substantially (12.8% +/- 13.7%). Interscan, intrasubject reproducibility was better for the automated than the manual method (2.9% +/- 3.2% vs 12.4% +/- 16.2% volume difference, P = .02). CONCLUSION: The automated algorithm appeared more reproducible, which renders it superior to the manual method for longitudinal studies.

Neural correlates of attention and working memory deficits in HIV patients.

OBJECTIVES: To evaluate the neural correlates of attention and working memory deficits in patients with HIV-1. METHOD: fMRI was used to evaluate brain activity in 11 patients with HIV and 11 age-, sex-, education-, and handedness-matched seronegative subjects, while performing a battery of tasks that required different levels of attention for working memory. RESULTS: Patients with HIV showed greater brain activation (blood oxygenation level dependent signal changes) in some regions compared with control subjects while performing the same tasks. For the simpler tasks, patients with HIV showed greater activation in the parietal regions. However, with more difficult tasks, patients with HIV showed greater activation additionally in the frontal lobes. Reaction times during these tasks were slower but accuracy was similar in the patients with HIV compared with control subjects. CONCLUSION: Injury to the neural substrate caused by HIV infection may necessitate greater attentional modulation of the neural circuits, hence a greater use of the brain reserve; additional activation of the frontal lobes is required to perform the more complex tasks. The task-dependent increased frontal activation in patients with HIV suggests that the neural correlate of attentional deficits may be excessive attentional modulation as a result of frontostriatal brain injury.

Computational modelling of visual attention.

Five important trends have emerged from recent work on computational models of focal visual attention that emphasize the bottom-up, image-based control of attentional deployment. First, the perceptual saliency of stimuli critically depends on the surrounding context. Second, a unique 'saliency map' that topographically encodes for stimulus conspicuity over the visual scene has proved to be an efficient and plausible bottom-up control strategy. Third, inhibition of return, the process by which the currently attended location is prevented from being attended again, is a crucial element of attentional deployment. Fourth, attention and eye movements tightly interplay, posing computational challenges with respect to the coordinate system used to control attention. And last, scene understanding and object recognition strongly constrain the selection of attended locations. Insights from these five key areas provide a framework for a computational and neurobiological understanding of visual attention.

Automatic scan prescription for brain MRI.

Diagnostic brain MRI scans are usually performed by trained medical technologists who manually prescribe the position and orientation of a scanning volume. In this study, a fully automatic computer algorithm is described which compensates for variable patient positioning and acquires brain MRI scans in a predefined reference orientation. The method involves acquiring a rapid water-only pilot scan, segmenting the brain surface, and matching it to a reference surface. The inverse matching transformation is then used to adapt a geometric description of the desired scanning volume, defined relative to the reference surface, to the current patient. Both pilot scan and processing are performed within 30 sec. The method was tested in 25 subjects, and consistently recovered orientation differences between the reference and each subject to within +/-5 degrees. Compared to manual prescription, automatic scan prescription promises many potential benefits, including reduced scan times, reproducible scan orientations along anatomically preferable orientations, and better reproducibility for longitudinal studies. Magn Reson Med 45:486-494, 2001.

Changes in cerebral metabolism are detected prior to perfusion changes in early HIV-CMC: A coregistered (1)H MRS and SPECT study.

Human immunodeficiency virus-cognitive motor complex (HIV-CMC), a common complication of the acquired immunodeficiency syndrome (AIDS), is characterized by progressive cognitive impairment and motor dysfunction. Functional imaging methods, such as single-photon emission computed tomography (SPECT) and proton magnetic resonance spectroscopy ((1)H-MRS), have been applied to assess the severity of brain injury. However, it is unclear which of these two methods is more sensitive in detecting brain abnormalities in patients with early HIV-CMC. Twenty-four HIV-CMC patients were compared with 34 healthy subjects; each had quantitative SPECT ((133)Xenon-calibrated (99m)Tc-HMPAO) and quantitative (1)H-MRS. Both modalities were co-registered in order to assess regional cerebral blood flow (rCBF) and metabolite concentrations within the same voxel of interest in four brain regions (midfrontal and midparietal gray matter, temporoparietal white matter, and basal ganglia). On SPECT, only the temporoparietal white matter showed a trend for decreased rCBF in HIV-CMC patients (-13%, P = 0.06). On MRS, HIV-CMC patients showed significantly reduced creatine concentration in the basal ganglia (-8%, P = 0.008), as well as increased myoinositol concentrations in the basal ganglia (+25%, P = 0.01) and the temporoparietal white matter (+18%, P = 0.08). There was no significant correlation between SPECT and MRS variables in the patients in any region. (1)H MRS showed abnormal neurochemistry in the basal ganglia, whereas rCBF on SPECT was normal in the same region. This finding suggests that metabolite concentrations on (1)H MRS are better surrogate markers than rCBF measurements with SPECT for the evaluation of brain injury in early HIV-CMC. J. Magn. Reson. Imaging 2000;12:859-865.

Revisiting spatial vision: toward a unifying model.

We report contrast detection, contrast increment, contrast masking, orientation discrimination, and spatial frequency discrimination thresholds for spatially localized stimuli at 4 degrees of eccentricity. Our stimulus geometry emphasizes interactions among overlapping visual filters and differs from that used in previous threshold measurements, which also admits interactions among distant filters. We quantitatively account for all measurements by simulating a small population of overlapping visual filters interacting through divisive inhibition. We depart from previous models of this kind in the parameters of divisive inhibition and in using a statistically efficient decision stage based on Fisher information. The success of this unified account suggests that, contrary to Bowne [Vision Res. 30, 449 (1990)], spatial vision thresholds reflect a single level of processing, perhaps as early as primary visual cortex.

A saliency-based search mechanism for overt and covert shifts of visual attention.

Most models of visual search, whether involving overt eye movements or covert shifts of attention, are based on the concept of a saliency map, that is, an explicit two-dimensional map that encodes the saliency or conspicuity of objects in the visual environment. Competition among neurons in this map gives rise to a single winning location that corresponds to the next attended target. Inhibiting this location automatically allows the system to attend to the next most salient location. We describe a detailed computer implementation of such a scheme, focusing on the problem of combining information across modalities, here orientation, intensity and color information, in a purely stimulus-driven manner. The model is applied to common psychophysical stimuli as well as to a very demanding visual search task. Its successful performance is used to address the extent to which the primate visual system carries out visual search via one or more such saliency maps and how this can be tested.

Effect of ecstasy [3,4-methylenedioxymethamphetamine (MDMA)] on cerebral blood flow: a co-registered SPECT and MRI study.

3,4-methylenedioxymethamphetamine (MDMA), an illicit recreational drug, damages serotonergic nerve endings. Since the cerebrovasculature is regulated partly by the serotonergic system, MDMA may affect cerebral blood flow (CBF) in humans. We evaluated 21 abstinent recreational MDMA users and 21 age- and gender-matched healthy subjects with brain SPECT and MRI. Ten of the MDMA subjects also had repeat SPECT and MRI after receiving two doses of MDMA. Abstinent MDMA users showed no significantly different global or regional CBF (rCBF) compared to the control subjects. However, within 3 weeks after MDMA administration, rCBF remained decreased in the visual cortex, the caudate, the superior parietal and dorsolateral frontal regions compared to baseline rCBF. The decreased rCBF tended to be more pronounced in subjects who received the higher dosage of MDMA. Two subjects who were scanned at 2-3 months after MDMA administration showed increased rather than decreased rCBF. Low-dose recreational MDMA use does not cause detectable persistent rCBF changes in humans. The lack of long-term rCBF changes may be due to a non-significant effect of serotonergic deficits on rCBF, or regeneration of serotonergic nerve terminals. The subacute decrease in rCBF after MDMA administration may be due to the direct effect of MDMA on the serotonergic system or the indirect effects of its metabolites on the dopaminergic system; the preliminary data suggest these effects may be transient.

Simultaneous correction for interscan patient motion and geometric distortions in echoplanar imaging.

A method is presented for simultaneous correction of linear geometric distortions and interscan patient motion in echoplanar imaging (EPI). The technique does not require the acquisition of specialized scans other than high-resolution magnetic resonance images. The method is based on a generalized surface-based coregistration algorithm, which accounts for a complete 3-dimensional affine transformation, i.e., rotations, translations, scaling, and shearing, between two volumetric image data sets. Any minimally distorted high-resolution scan may serve as a reference data set, to which the EPI data set is matched. The algorithmic accuracy was assessed using simulated data sets with known affine distortions. The deviation of the parameters determined by the coregistration program from the true values typically was 1% or less. Precise alignment of functional and anatomic information will be important for many future clinical applications.

Comparison of static and dynamic MRI techniques for the measurement of regional cerebral blood volume.

Two different acquisition and processing strategies to determine the regional cerebral blood volume (rCBV) with magnetic resonance imaging (MRI) are compared. The first method is based on the acquisition of the signal time course during a bolus administration of a contrast agent (dynamic method). The second method evaluates signal changes before and after the contrast agent injection (static method), assuming the contrast agent remains primarily intravascular in the brain after the first pass. Both methods were applied to the same data sets, acquired with either echoplanar imaging (EPI, n = 18) or fast low-angle shot (FLASH, n = 28) techniques. A voxel-by-voxel correlation between the static and dynamic method yielded a correlation coefficient of 0.76 +/- 0.06 for the EPI and 0.71 +/- 0.10 for the FLASH measurements. The static method was less sensitive and showed higher standard deviations for rCBV than the dynamic method. With the development of truly intravascular contrast agents, the static perfusion MRI method, which can be performed with higher signal-to-noise ratio and higher spatial resolution, may become an alternative to ultra-fast MRI for measuring rCBV.

Attention activates winner-take-all competition among visual filters.

Shifting attention away from a visual stimulus reduces, but does not abolish, visual discrimination performance. This residual vision with 'poor' attention can be compared to normal vision with 'full' attention to reveal how attention alters visual perception. We report large differences between residual and normal visual thresholds for discriminating the orientation or spatial frequency of simple patterns, and smaller differences for discriminating contrast. A computational model, in which attention activates a winner-take-all competition among overlapping visual filters, quantitatively accounts for all observations. Our model predicts that the effects of attention on visual cortical neurons include increased contrast gain as well as sharper tuning to orientation and spatial frequency.

Correlation of regional cerebral blood flow from perfusion MRI and spect in normal subjects.

The objective of this study was to determine the relationship in regional cerebral blood flow (rCBF) as measured with perfusion magnetic resonance imaging (pMRI) and single photon emission computer tomography (SPECT). rCBF was determined in 26 healthy subjects with pMRI and SPECT. After co-registration of pMRI with SPECT, rCBF was determined in 10 brain regions relative to the whole slice value. pMRI was evaluated with and without elimination of large vessels. rCBF from pMRI correlates significantly with rCBF from SPECT (r = 0.69 with and r = 0.59 without elimination of large vessels; p < 0.0001 for both). Elimination of large vessels reduced the interindividual variance of the pMRI measurements in most regions. rCBF from pMRI shows good correlation with rCBF from SPECT. Because pMRI is sensitive to flow in large vessels while SPECT is not, elimination of large vessels in pMRI reduces the interindividual variability of pMRI and improves the-correlation between the two methods. pMRI is a reliable noninvasive method for rCBF measurements.

Midline cerebral morphometry distinguishes frontotemporal dementia and Alzheimer's disease.

We investigated and contrasted midline cerebral structures in frontotemporal dementia (FTD) and Alzheimer's disease (AD). FTD and AD may be difficult to distinguish clinically. FTD typically affects frontal and anterior temporal regions, whereas AD tends to involve more posterior temporal and parietal areas. We hypothesized that disease-specific cerebral alterations would be differentially reflected in corresponding regions of the corpus callosum (CC), pericallosal CSF space (PCS), or their ratio (CC:PCS). Regions-of-interest (ROIs) from midsagittal MRIs in 17 AD, 16 FTD, and 12 elderly control (EC) subjects were analyzed. ROIs were divided into four regions using an anatomic landmark-based computer algorithm and were adjusted for head size variation. FTD subjects had a much smaller anterior CC region and significantly larger PCS area, particularly in anterior regions. AD and EC subjects did not differ significantly in any total or regional ROI measure. Total and anterior CC:PCS ratios were markedly lower in FTD patients. Across groups, total CC:PCS correlated significantly with midsagittal cerebral area and was similarly associated with Mini-Mental State Examination score. Anterior CC (AD) and PCS (FTD) regions exhibited disease-specific relationships to these variables. A discriminant model using two ROI variables correctly classified 91% of AD and FTD patients, comparing favorably with blind clinical MRI diagnostic ratings. Midline cerebral structural alterations reflect differential patterns of cerebral degeneration in AD and FTD, yielding morphometric indices that may facilitate the study of brain-behavior relationships and differential diagnosis of dementia.

Robust multimodality registration for brain mapping.

We present a robust intrasubject registration method for the synergistic use of multiple neuroimaging modalities, with applications to magnetic resonance imaging (MRI), functional MRI, perfusion MRI, MR spectroscopy, and single-photon emission computed tomography (SPECT). This method allows user-friendly processing of difficult examinations (low spatial resolution, advanced pathology, motion during acquisition, and large areas of focal activation). Registration of three-dimensional (3D) brain scans is initially estimated by first-order moment matching, followed by iterative anisotrophic chamfer matching of brain surfaces. Automatic brain surface extraction is performed in all imaging modalities. A new generalized distance definition and new specific methodologies allow registration of scans that cover only a limited range of brain surface. A new semiautomated supervision scheme allows fast and intuitive corrections of possible false automatic registration results. The accuracy of the MRI/SPECT anatomical-functional correspondence obtained was evaluated using simulations and two difficult clinical populations (tumors and degenerative brain disorders). The average discrimination capability of SPECT (12.4 mm in-plane resolution, 20 mm slice thickness) was found to be better than 5 mm after registration with MRI (5 mm slice thickness). Registration accuracy was always better than imaging resolution. Complete 3D MRI and SPECT registration time ranged between 6-11 min, in which surface matching represented 2-3 min. No registration failure occurred. In conclusion, the application of several new image processing techniques allowed efficient and robust registration. Hum. Brain Mapping 5:3-17, 1997. (c) 1997 Wiley-Liss, Inc.

Improved 3D correction for partial volume effects in brain SPECT.

An improved method for correction of partial volume effects (PVE) in brain SPECT is proposed. It is fully three-dimensional, does not require particular patient positioning, and works with scans only partially covering the brain. The location of functionally inactive brain regions (primarily cerebrospinal fluid) is extracted from high-resolution MRI. An automatic 3D registration algorithm then determines the geometric transformation between MRI and SPECT. Correction consists of: 1) counting the volumetric active/inactive ratio in each volume element of the functional scan using the measured SPECT point spread function; 2) correcting the functional measures according to these ratios; 3) fusing functional and anatomical information at the resolution of MRI. Quantitative validation was performed using a phantom containing a test region in which multiple parallel acrylic plates thinner than SPECT resolution created high PVE, as well as a large reference region not suffering from PVE. Reference activity was recovered in the test region with an accuracy of 1-3%. The method was applied to clinical images demonstrating a combination of hypoperfusion and cortical atrophy. The composite anatomical-functional corrected images, in which the main sulci are visible, yield better differentiation between decreased function and focal atrophy.

Cerebral blood flow correlates of apathy in Alzheimer disease.

BACKGROUND: Apathy is a pervasive noncognitive neuropsychiatric disturbance in Alzheimer disease, which causes significant caregiver distress. The neuroanatomical substrate of apathy is not well understood. OBJECTIVE: To study the relationship between regional cerebral blood flow and the presence and severity of the personality disturbance, apathy, in individuals with Alzheimer disease. DESIGN: Analysis of the relationship between regional cerebral blood flow as measured by single photon emission computed tomography and severity of apathy as measured by the Neuropsychiatric Inventory using an analysis of variance design. We examined regional cerebral perfusion alterations as measured by xenon 133Xecalibrated technetium Tc 99m hexamethyl-propyleneamine-oxime single photon emission computed tomography in relation to the presence and severity of apathy. SETTING: The neurology clinics of the University of California, Los Angeles, UCLA School of Medicine, and Harbor-UCLA Medical Center. PARTICIPANTS: Thirty-one community-dwelling patients fulfilling National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association diagnostic criteria for probable Alzheimer disease who had a single photon computed tomographic scan performed within 3 months of administration of the Neuropsychiatric Inventory. RESULTS: The presence of apathy was associated with more severe prefrontal and anterior temporal dysfunction. These regional cerebral perfusion relationships with apathy were independent of cognitive decline except in the dorsolateral prefrontal cortex. CONCLUSIONS: These results demonstrate the association of apathetic syndromes with prefrontal and anterior temporal regional brain dysfunction and are consistent with similar findings previously reported in other disorders.

Neural basis of confabulation.

We present a case of acute alcohol-induced Korsakoff amnesia. A severe amnestic-confabulatory syndrome characterized the early clinical status. The initial neuropsychological tests demonstrated severe learning deficits plus impaired performance on many, but not all, tests of frontal lobe function. Single-photon emission CT (SPECT) at this stage showed hypoperfusion in the orbital and medical frontal regions and the medial diencephalic area. Four months later, the patient's amnesia remained but there was no confabulation. Repeat neuropsychological tests confirmed an ongoing severe amnesia, but performance on the frontal lobe tests now was normal. Repeat SPECT showed a return to normal perfusion in the frontal brain areas but little improvement in the medial diencephalic region. These findings along with data from the clinical literature suggest that confabulation results from dysfunction of orbital and a medial frontal cortex.