Advanced

Ocular Dominance Columns: Retinotopy and Detection

Ikdal, Erik Gaustad LU (2016) BMEM01 20162
Department of Biomedical Engineering
Abstract
This report discusses two research questions concerning ocular dominance columns, a cortical structure present in the primary visual cortex which predominately respond to either left or right eye stimuli. The first research question addresses the problem of retinotopy and ocular dominance columns. Retinotopy is the mapping of visual stimuli in the retina to a location in the visual cortex. Currently, most retinotopic models are continuous and monocular, and can therefore not take into account binocular structures such as ocular dominance columns. In this report, this is rectified by combining an artificial neural network called a self organising map with a retinotopic mapping function. This model creates a fully binocular model which not... (More)
This report discusses two research questions concerning ocular dominance columns, a cortical structure present in the primary visual cortex which predominately respond to either left or right eye stimuli. The first research question addresses the problem of retinotopy and ocular dominance columns. Retinotopy is the mapping of visual stimuli in the retina to a location in the visual cortex. Currently, most retinotopic models are continuous and monocular, and can therefore not take into account binocular structures such as ocular dominance columns. In this report, this is rectified by combining an artificial neural network called a self organising map with a retinotopic mapping function. This model creates a fully binocular model which not only produces strikingly similar patterns to that of ocular dominance columns, but simulations show that the retinotopic behaviour is similar to previous physiological experiments.

The second research question investigates the possibility of detecting ocular dominance columns using functional Magnetic Resonance Imaging (fMRI). Experimental evidence suggests that fMRI is able to detect ocular dominance columns despite it's structure being at the very limit of fMRI spatial resolution. In this work, a spatiotemporal model derived from physiology is used to simulate the effects of ocular dominance column hemodynamics as measured by an fMRI. Using a mass univariate approach to analyse the simulated signal, results showed that although possible to observe ocular dominance columns at certain location, much of the fine scale detail is lost. (Less)
Popular Abstract
Detection of Ocular Dominance Columns

The primary visual cortex or V1, is a centre for visual stimuli processing in the brain. Extensive research trying to understand the function and anatomy of the V1 is of paramount importance when trying to describe how the brain processes visual information. Two pioneers in the field, David Hubel and Torsten Wiesel studied the V1 of cats during the 1950's and 1960's. Using an electrode while presenting visual stimuli, they found that at certain areas of the V1 neurons only respond to certain types of visual stimuli. This was in agreement with studies conducted by Vernon Mountcastle, who studied the somatosensory cortex, a location of the brain which is responsible for the sense of touch.... (More)
Detection of Ocular Dominance Columns

The primary visual cortex or V1, is a centre for visual stimuli processing in the brain. Extensive research trying to understand the function and anatomy of the V1 is of paramount importance when trying to describe how the brain processes visual information. Two pioneers in the field, David Hubel and Torsten Wiesel studied the V1 of cats during the 1950's and 1960's. Using an electrode while presenting visual stimuli, they found that at certain areas of the V1 neurons only respond to certain types of visual stimuli. This was in agreement with studies conducted by Vernon Mountcastle, who studied the somatosensory cortex, a location of the brain which is responsible for the sense of touch. Mountcastle called these type of clustering neurons columns, thus the idea of columns in the V1 was born.

Hubel and Wiesel found two types of columns: Ocular Dominance (OD) columns which respond either to left or right eye stimuli and Orientation Columns which respond to the orientation of the presented stimuli. The structure of OD columns are often described as a zebra like pattern on the V1 where the black and white stripes represent either left or right eye neurons. Although OD column structure has been measured in several species including humans, their function still largely remain a mystery.

Since Hubels and Wiesels discovery different techniques for revealing OD structures has been invented, the problem with these techniques is that they are highly invasive and are very impractical if the studies are to be conducted on many individuals. fMRI however is a non-invasive medical imaging technique with the possibility of studying sub-millimetre structures such as OD columns. Although several articles have reported OD columns using fMRI, it is difficult to quantify how well this technique performs since we are actually not able to physically peek inside the brain.

mVibe is an fMRI simulation environment, which is designed to mimic the measurements of an fMRI. A realistic simulation environment such as mVibe enables the possibility to see how well OD detection using fMRI performs. The two main research question of the thesis was to purpose a model to extend mVibe to take into account binocular OD columns and to simulate the fMRI responds using the biophysical models from mVibe.

The proposed binocular model showed great similarities to real measurements of the behaviour of OD columns. The model also purposes a few predictions that could possibly be used to test the assumptions of the model. The simulated fMRI activity showed that although possible to measure OD columns, the overall structure will tend to cluster together making it difficult to image OD columns at certain locations.

Using fMRI to study small scale structures such as cortical columns not only opens the possibility to better understand their function, but is also more ethically viable for humans and animals. The obtained results might also be used to create other techniques of analysis which are more sensitive to OD column like structures. (Less)
Please use this url to cite or link to this publication:
author
Ikdal, Erik Gaustad LU
supervisor
organization
course
BMEM01 20162
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Ocular Dominance Columns Retinotopy fMRI Self Organizing Maps
language
English
additional info
2016-17
id
8893935
date added to LUP
2016-10-24 16:04:23
date last changed
2016-10-28 09:34:32
@misc{8893935,
  abstract     = {This report discusses two research questions concerning ocular dominance columns, a cortical structure present in the primary visual cortex which predominately respond to either left or right eye stimuli. The first research question addresses the problem of retinotopy and ocular dominance columns. Retinotopy is the mapping of visual stimuli in the retina to a location in the visual cortex. Currently, most retinotopic models are continuous and monocular, and can therefore not take into account binocular structures such as ocular dominance columns. In this report, this is rectified by combining an artificial neural network called a self organising map with a retinotopic mapping function. This model creates a fully binocular model which not only produces strikingly similar patterns to that of ocular dominance columns, but simulations show that the retinotopic behaviour is similar to previous physiological experiments. 

The second research question investigates the possibility of detecting ocular dominance columns using functional Magnetic Resonance Imaging (fMRI). Experimental evidence suggests that fMRI is able to detect ocular dominance columns despite it's structure being at the very limit of fMRI spatial resolution. In this work, a spatiotemporal model derived from physiology is used to simulate the effects of ocular dominance column hemodynamics as measured by an fMRI. Using a mass univariate approach to analyse the simulated signal, results showed that although possible to observe ocular dominance columns at certain location, much of the fine scale detail is lost.},
  author       = {Ikdal, Erik Gaustad},
  keyword      = {Ocular Dominance Columns Retinotopy fMRI Self Organizing Maps},
  language     = {eng},
  note         = {Student Paper},
  title        = {Ocular Dominance Columns: Retinotopy and Detection},
  year         = {2016},
}