The purpose of the project is to carry out an interdisciplinary analysis of motion processing in the visual system, which will combine anatomical, physiological (single-cell recording and evoked potentials), psychological, computational, neural imaging and clinical approaches.
Anatomical mapping will be carried out of point to point topography and laminar and cellular targets of connections between cortical areas; the topographic relationships will be determined within regions showing motion specific responses as measured by physiological unit recording. The spatial dimensions of the anatomical maps will be compared to the spatio-temporal constraints of psycophysical and physiological measures of motion detection and direction specificity to help generate data for models of motion and direction detection (Lund, Burr, Morgan, Fahle: with colleagues Levitt, Solomon, Morrone).
Anatomical and physiological characteristics of direction selective neurons in layer 6 of V1 will be explored in terms of the role of cortical feedback to LGN (including possible suppression of the motion signal in conjunction with subcortical inputs from the magnocellular pathway during saccades) and in terms of recurrent enervation of layer 4 of V1 and outflow to MT and V2 ( Lund with Levitt, Asi and Sillito; Burr, Morrone).
Construction and psycophysical testing of computer models of the biological substrate for motion direction selectivity (Lund -with colleague K. Obermayer, Morgan, Morrone, Fahle).
Psycophysical experiments investigating the input from motion mechanisms to binocular stereopsis (Morgan, Burr, Fahle).
Psychophysical experiments investigating the higher level integration of the Motion Signal and motion deblurring (Burr, Fahle, Morgan).
Brain Imaging During Motion Tasks (Fahle, Paakonen, Burr, Morgan).
Within the area of Vision, the processing of the motion signal from the retina is one of the fundamental functions of the visual system. The motion signal is important in detecting moving objects against their stationary surroundings; in helping to build up the three-dimensional structure of the world from the motion parallax and stereo-motion systems; and for controlling locomotion and posture. It is also important that the motion signal should be controlled by feedback, during saccadic eye movements and nystagmus. Abnormalities of motion processing can be an early sign of neurodegenerative diseases such as glaucoma and multiple sclerosis, and there are probably many abnormalities that have hitherto gone unrecognized, for example, in stereo-motion.
There is a major input to motion perception through the achromatic, low-resolution "M" pathway, originating in alpha ganglion cells in the retina and proceeding through magnocellular layers in LGN, and layers 4c-alpha then 4b of V1. Relays then pass to cortical area MT (V5), which is thought to play a major role in the spatial integration of the motion signal needed to analyse the optic flow field. Relays also leave layer 4b to terminate in areas V2 and V3. Large-scale integration of the low-level motion signal is necessary because local signals are ambiguous (Hildreth, 1983). Lesions of the magnocellular layers of the LGN cause deficits in detection of temporally-transient stimuli (Schiller, Logothetis & Charles, 1990), and there is clinical evidence that links "motion blindness" to damage to MT. However, the view that the retina to geniculate, geniculate to area V1 M-pathway is solely responsible for motion analysis is certainly too simple. There is substantial recovery of direction selectivity after removal of area V1 (Rodman et al.,1989); convergence of M and P pathways in V1 (Yoshioka, Levitt & Lund, 1994); and, contrary to earlier reports, interaction between chromatic and achromatic inputs to motion (Cavanagh & Favreau, 1990; Morgan & Ingle, 1994, Gegenfurtner et al., 1996). The interaction between M and P pathways in motion requires further investigation, one possibility is that the M pathway is involved mainly with high velocities/low spatial frequencies/high temporal frequencies, the P pathway being involved in additional cues for motion such as hue and slow, fine-grain stimuli. A major theme running through the project is the role of M and P pathways in motion processing.
Our objective is to form a strong inter-disciplinary group of workers in motion, comprising anatomists, physiologists, psychophysicists, medical physicists (brain imaging), and clinical workers, and to institute collaborative research projects within this group. We aim to have twice-yearly meetings of all members of the group as well as specific laboratory visits. Our scientific objectives are (1) to model the early stages of motion selectivity (2) to understand the interactions underlying the receptive field properties of directionally-selective neurones (3) to understand the importance of feedback in elaborating the motion signal, and in suppressing it during eye movements (4) to investigate specific defects of motion in clinical populations.
Funding SchemeCSC - Cost-sharing contracts