A New Method of Measuring Retinal Vessel Diameters from Fundus Photographs Based on Retinal Structure and Light Reflection

<p>We propose a new method and software for measuring blood vessel diameters from fundus photographs. In this method the edge of the blood vessel is the two minimum points in the luminance profile and diameter is distance between minimum points. We noticed the relation between W-shape of luminance profile at blood vessel transversal line and cylindrical shape, and analyzed the validity of this method using the diffuse reflection model of a cylinder shape based on the optics. The simulation image by the diffuse reflection model of the cylinder was similar to that of the blood vessel images. A newly developed software using the minimum points can measure repeated measurement in a series of photographs at the same point of the same person. The data of the repeated measurement of each person showed stability, and the coefficient of variation showed no affection of different diameter and arteriovenous distinction. Hence, we believe that the minimum points are valid as blood vessel edges. The software that measures these blood vessel edges has an extremely simple measurement procedure, possesses few errors, and is easy to automate. This method is relevant because the central reflex, which has not been utilized previously, optically represents the cylindrical three-dimensional structure of blood vessels.</p><div><div> </div> </div>

A New Method of Measuring Retinal Vessel Diameters from Fundus Photographs Based on Retinal Structure and Light Reflection

<p>We propose a new method and software for measuring blood vessel diameters from fundus photographs. In this method the edge of the blood vessel is the two minimum points in the luminance profile and diameter is distance between minimum points. We noticed the relation between W-shape of luminance profile at blood vessel transversal line and cylindrical shape, and analyzed the validity of this method using the diffuse reflection model of a cylinder shape based on the optics. The simulation image by the diffuse reflection model of the cylinder was similar to that of the blood vessel images. A newly developed software using the minimum points can measure repeated measurement in a series of photographs at the same point of the same person. The data of the repeated measurement of each person showed stability, and the coefficient of variation showed no affection of different diameter and arteriovenous distinction. Hence, we believe that the minimum points are valid as blood vessel edges. The software that measures these blood vessel edges has an extremely simple measurement procedure, possesses few errors, and is easy to automate. This method is relevant because the central reflex, which has not been utilized previously, optically represents the cylindrical three-dimensional structure of blood vessels.</p><div><div> </div> </div>

Complementary roles of two excitatory pathways in retinal directional selectivity

The two major excitatory synapses onto ON–OFF directionally selective (DS) ganglion cells of the rabbit retina appear to be nicotinic cholinergic and NMDA glutamatergic. Blockade of either of these synapses with antagonists does not eliminate directional selectivity. This suggests that these synapses may have complementary roles in the computation of the direction of motion. To test this hypothesis, quantitative features of the DS cell excitatory pathways were determined by collecting responses, under nicotinic and/or NMDA blockade, to a sweeping bar, hyperacute apparent motions, or a drifting sinusoidal grating. Sweeping bar responses were reduced, but directional selectivity not eliminated, by blockade of either excitatory path, as previously shown (Cohen & Miller, 1995; Kittila & Massey, 1997). However, residual responses under combined blockades were not statistically significantly DS. NMDA blockade reduced responses more than nicotinic blockade for each protocol, and shifted hyperacute motion thresholds to higher values. This supported the notion that glutamate provides the main excitatory drive to DS cells, that is, the one responsible for contrast sensitivity. In turn, nicotinic, but not NMDA blockade eliminated directional selectivity to a drifting low spatial-frequency sinusoidal grating in these cells. This suggested that acetylcholine (ACh) is the main excitatory input with regards to directional selectivity for some textured stimuli, that is, those with multiple peaks in their spatial luminance profile. Moreover, nicotinic blockade raised the low temporal-frequency cutoff of the grating responses, consistent with the proposal that preferred-direction facilitation, which is temporally sustained, is dependent on the cholinergic input. These different properties of the NMDA and nicotinic pathways are consistent with a recently proposed two-asymmetric-pathways model of directional selectivity.

Selective Attention to Specific Location Cues: The Peak and Center of a Patch are Equally Accessible as Location Cues

Asymmetric patterns have several spatially distinct cues for spatial localization. These cues include the peak of the luminance distribution, the centroid of the contrast distribution, zero-crossings in the second derivative of the luminance profile, and the midpoint of the visible area. If these cues are represented as primitives in the visual system, the observer should be able to access them at will. To examine whether observers can selectively attend to particular cues, we measured perceived alignment for an asymmetric pattern with two distinct instructions: “align the peak”, and “align the center”. We found that observers could align the patterns in accord with the instructions with identical precision, suggesting that the peak and the center cues were equally accessible by the observer. We conclude that multiple localization cues are represented in and can be selectively accessed by the visual system.

Visually Evoked Magnetic Field Induced by Ring Motion Aftereffect

Motion aftereffect (MAE) is a negative aftereffect caused by prolonged viewing of visual motion: after gazing at a moving grating for a while, a stationary image will appear to move in the opposite direction (Ashida and Osaka, 1995 Vision Research35 1825). Evoked magnetic field (magnetoencephalogram: MEG) was measured when a human subject observing ring MAE in which concentric circles appear to contract continuously after viewing continuously expanding rings. The diameter of the stimulus was 20 deg with fixation point in the centre. The magnetic evoked field (80 averagings at a latency of 190 ms) was measured from 37 points over the occipital and parietal areas (Magnes SQUID biomagnetometer, BTi) while the subject was observing stationary rings after an adaptation period of 2 s at low spatial frequency (4 cycles deg−1, 4 Hz). The luminance profile was sinusoidally changed across rings. MRI image fitting (sagittal, coronal, and axial view) for each of four subjects, and dipole estimates obtained for equal magnetic field contours (with value of goodness of fit greater than 0.98) from the right brain hemisphere suggest that the main loci subserving MAE lie in the surrounding region over the lateral occipitotemporal areas in the human brain, close to area MT. This is in good agreement with another study with fMRI-based MAE measures [Tootell et al, 1995 Nature (London)375 139] in which a clear increase in activity in these areas was observed when subjects viewed MAE.