• Home
• People
• Research
• Publications
• Software
• Demos
• Internal

The research spans an interdisciplinary cross-section of engineering, psychology, and neuroscience. In the fields of perceptual psychology and systems/cognitive neuroscience, we have worked on computational models of neuronal processing in the visual system, psychophysical (perceptual psychology) measurements of human vision, and neuroimaging. In the fields of image processing, computer vision, and computer graphics, we have worked on motion estimation and image registration, wavelet image representations, anisotropic diffusion (edge-preserving noise reduction), image fidelity metrics (for evaluating image data compression algorithms), texture analysis/synthesis and scientific visualization.

Neuroimaging

One current focus of the research in our lab is to use functional magnetic resonance imaging (fMRI) to quantitatively investigate the relationship between brain and behavior. The vast majority of neuroimaging experiments from other labs around the world have focused on which parts of the brain are involved in a particular cognitive or perceptual task. Although this has been an important first step, perception and cognition depend not only on which brain areas are active, but also on how neuronal activity within each of those areas varies over space and time. We are using fMRI to measure the timing and amplitude of brain activity, for testing computational theories of the neural processing underlying cognition and perception. Part of my own excitement about this work is that it brings together my engineering training with my interest in neuroscience, as we routinely develop new image processing and computer vision algorithms for analyzing our functional and structural MRI data. We are using fMRI to study visual awareness, visual pattern detection/discrimination, visual motion perception, stereo depth perception, attention, working memory, the control of eye and hand movements, and neural processing of complex audio-visual and emotional experiences (movies, music, narrative). See below for a complete list.

Neuroimaging, particularly functional magnetic resonance imaging, has revolutionized neuroscience over the past decade. Along with the revolution in neuroimaging, a new field of neuroethics has evolved. Neuroethics is the study of the ethical, legal and social questions arising when scientific findings about the brain are carried into medical practice, legal interpretations, and health and social policy. There are a host of ethical concerns including privacy (reading someone's mind with fMRI) and culpability (should someone be held responsible for a crime if an fMRI measurement can show that there is something wrong with their brain), etc. The Dana Foundation is a good place to start for finding information about neuroethics. A pressing example of neuroethics concerns lie detection. Two companies (Cephos and NoLieMRI) have announced new lie detection technologies based on fMRI. I participated in an ACLU press briefing on this topic entitled "Mining the Mind", which can be downloaded from the ACLU web site. The ACLU has also issued a request under the Freedom of Information Act for information about the government use of brain scanners in interrogations (see ACLU press release). The issue has been covered by articles in Nature and Science, as well as USA Today, the SF Chronicle, the New York Times, The New Yorker, and a number of other newspapers and magazines.

Computational Neuroscience

Another current focus of the research in our lab is to develop computational theories of brain function. A variety of anatomical, physiological, and behavioral evidence suggests that the brain performs computations using modules that are repeated across species, brain areas, and modalities. What are these canonical modules, and how can we elucidate their underlying circuitry and mechanisms? A classical example of canonical computation is the linear receptive field. It has been found to be a powerful description of neuronal responses in the visual system including primary visual cortex and area MT, in somatosensory cortex, and in auditory cortex. A second example of canonical computation is soft-thresholding of noisy signals. The conversion of input currents into output firing rates introduces a thresholding stage. This threshold sets the operating point of visual cortex, allows neurons to have invariant tuning curves, and to effectively amplify the variability of their responses. A third example of canonical computation is divisive gain control which has been found to be a key computation in many neural systems.

We are focusing on a model of canonical neural computation, called "the normalization model", that encompasses the linear receptive field, soft-thresholding, and divisive suppression. The normalization model has been proposed to explain the physiology of primary visual cortex (V1), and cortical visual area MT. We have also extended the model to account for a wide variety of modulatory effects of attention on responses in visual cortex. See below for a complete list of publications on these and related topics.

For further information about canonical neural computation, see Canonical Neural Computation: A Summary and a Roadmap, from a workshop that was held on the topic.

Our research is funded primarily by the National Eye Institute and the National Institute of Mental Health. The fMRI data are acquired at the NYU Center for Brain Imaging.

David Heeger, Professor of Psychology and Neural Science, New York University