Monte Carlo estimation of spatial resolution of optical imaging in visual cortex and consequences for measurement of cortical functional architecture

Optical recording of brain activity uses electronic imaging to deduce hypothesized patterns of neural, metabolic, or hemodynamic activation. The spatial resolution of optical recording has not previously been quantitatively established. This work estimates the expected spatial point spread function and photon attenuation of optical recording in cortical tissue. Empirically measured in vivo scattering cross sections for 633 nanometer wavelength photons were used in a Monte Carlo simulation of accumulated photon scatter through different thicknesses of cortex, simulating the effect of different focal plane depths in a volume of tissue. A Gaussian beam decomposition of a standard (macroscope) lens system used in optical recording is used to model the point spread error due to the diffractive (wave optics) component of the instrumentation. The final system performance is estimated by a three-dimensional convolution of these two results, yielding the full-width-half-maximum (FWHM) of the point spread function and the photon attenuation, for values of focal plane and neural depth (i.e. laminar depth in the cortex) comparable to current experimental usage. The point spread function is used to interpret optical recording measurements of visual cortical orientation maps. Since cortical orientation response is characterized by a vector valued function defined over the cortical surface (i.e. a response amplitude and an orientation for each cortical location), its measurement is subject to a systematic spatial offset associated with the low-pass nature of optical recording. This causes the apparent location of the orientation singularities to be displaced from their true location, and can even cause neighboring singularities to “annihilate”, introducing error in the estimate of orientation singularity density and spacing. The results of this thesis provide quantitative estimates for the two major limitations of the use of optical recording in the measurement of cortical functional architecture. Strong photon attenuation tends to limit measurement to superficial laminae of the cortex, and low-pass filtering caused by the joint effects of optical depth of field and photon scatter causes a systematic shift in the location and density of orientation centers. These results are discussed in the context of quantitative and qualitative interpretation of contemporary orientation mapping experiments.