Functional MRI and Vision Science

Publications

• D. Ress and D. J. Heeger, Neuronal Correlates of Perception in Early Visual Cortex, Nature Neurosci. 6, 414 (2003)

• D. J. Heeger and D. Ress, What does fMRI Activity Tell Us about Neuronal Activity?, Nature Reviews Neuroscience 16, 142 (2002)

• D. Ress, B. T. Backus, D. J. Heeger, Activity in Primary Visual Cortex Predicts Performance in a Visual Detection Task, Nature Neurosci. 3, 940 (2000)

• A. C. Huk, D. Ress, and D. J. Heeger, Neuronal Basis of the Motion Aftereffect Reconsidered, Neuron 32, 161 (2001)
• G. H. Glover, T. Q. Li, D. Ress, Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn. Reson. Med. 44, 162 (2000)
  
  

Projects

· High resolution fMRI

· Real-time fMRI reconstruction and analysis

· Optimizing fMRI contrast

Functional contrast is established by the variable amounts of T₂^* decay manifested by neural tissue containing deoxygenated blood. We therefore use a long echo time, T_E, to enhance this contrast. If the functional image variations were entirely at low spatial frequencies, best contrast would be obtained for T_E = T₂^*. Because fMRI is often performed using long acquisitions, the T₂^* contrast variably affects different spatial frequencies as they are acquired. Given the spatial-frequency content of the functional perturbation, it is possible to calculate the optimal T_E. Click here for the detailed calculation.

· Hemodynamic modeling

The response to a brief (e.g., <2 s) interval of neural activity creates a lengthy response when measured using BOLD-contrast fMRI. We have found that this response, the BOLD hemodynamic impulse response (HIR), can be well modeled as the combination of two linear waveforms. One of these corresponds to the flow fluctuations induced by a brief vascular modulation in a hemodynamic network modeled by a 4-element windkessel, a lumped circuit consisting of two resistances, a compliant element, and an inertive element. This initially positive-going response is a damped sinusoid or sinh.

The second component is initially negative going. An early model attributed this component to a second vascular event within the parenchyma that would cause deoxygenated blood to stagnate. This model simply used a separate, unique 4-element windkessel as a model. This 5-parameter model was remarkably successful in describing a very large ensemble of measurements in several subjects; click here for details. A recent, more detailed model postulates that the negative-going component is the consequence of local diffusion from parenchymal capillaries. The corresponding response can be roughly simulated as a gamma function. This form of the model can also describe the same large dataset with similar efficacy. Details will appear soon.