|Year : 2017 | Volume
| Issue : 2 | Page : 66-67
Seeing vision with the eyes of math
V Srinivasa Chakravarthy
Department of Biotechnology, Bhupat and Jyoti Mehta School of Biosciences, Indian Institute of Technology Madras, Chennai, Tamil Nadu, India
|Date of Web Publication||10-Aug-2017|
V Srinivasa Chakravarthy
Department of Biotechnology, Bhupat and Jyoti Mehta School of Biosciences, Indian Institute of Technology Madras, Chennai, Tamil Nadu
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Chakravarthy V S. Seeing vision with the eyes of math. Kerala J Ophthalmol 2017;29:66-7
How do we see? This question, more than anything else, forms the essence of an ophthalmologist's profession. In an attempt to answer the question, the ophthalmologist studies the function of the eye in great detail. How the lens bends and focuses the incoming light, how the cornea aids the lens in this process of refraction, how the beam gets focused on the retina forming the retinal image, and how the retinal image, now converted into a series of electrical impulses, is transmitted, through the optic nerve, to the brain. Moreover, it is there that vision occurs, almost like a miracle, deep within the visual areas of the brain. The eye being the central object of study in ophthalmology, it is only natural that a massive attention is directed to that crucial organ of vision. However, the above approach to vision eclipses the deeper fact that a big part of seeing occurs, not in the eye, but beyond, as a result of neural information processing in the brain.
The question of “how we see?” reminds me of an interesting anecdote described in VS Ramachandran's engaging and illumining book – Phantoms in the Brain. Dr. Ramachandran recounts an interesting conversation he had with someone with a religious training and is a sort of a priest.
“What do you do for a living?” the priest asks.
“I study vision,” the neuroscientist replies.
“What is there to study vision? It is quite simple, isn't it?” the priest queries, now slightly perplexed.
“How do you think does vision work?” Dr. Ramachandran counters.
“Well, the image that falls on the retina is an inverted version of the world outside. The bundle of fibers from the eye to the brain twists around so that the image becomes upright when it reaches the brain. Then we see!”
The above conversation clearly puts the spotlight on the popular, lay understanding of seeing. Although every high school biology textbook has a diagram of eye and its parts, very little of how brain processes the visual image has entered general science education in India. However, the exciting branch of visual neuroscience had progressed tremendously worldwide. Since the path-breaking work of Hubel and Wiesel at Harvard in the late 1950s, our understanding of the organization of the visual areas of the brain has undergone an explosive growth. A new factor that had tremendously aided this growth is the application of mathematical methods to understand the workings of the visual system.
Thanks to the efforts of scientists from physics, mathematics, and engineering; there is currently emerging a novel perspective of brain function that is cast in the language of mathematics and computer science. The idea of using mathematics in biology, particularly neuroscience, may seem strange to some. This putative contradiction between biology and mathematics is also embodied in our school education in the form of choices given to a student: mathematics, physics, and chemistry or biology, physics, and chemistry. But it need not be so, if we recall that mathematics is the foundation of physical and chemical sciences. Engineering would be impossible without a solid mathematical theory of the relevant branch of science. Imagine sending a satellite into space aided by vague verbal theories of motion of an object in a gravitational field, or by graphical multi-colored descriptions of rocket propulsion!
One surprising realization that emerges out of a successful application of mathematical theory to biology is that behind all the intimidating complexity of biology systems, there is a beautiful simplicity and a deep unity. Thus, mathematical theories reveal the two faces, janus-like, of biology: one face of staggering complexity and the other of profound unity.
Application of mathematical theory to neuroscience in particular has nearly three decades of crowded history. These theoretical forays are greatly responsible to the revolution that is underway in contemporary neuroscience. It is now possible to describe brain function at multiple scales – molecular, cellular, local circuit level, module level, systems level, and so on. There are detailed mathematical models of how the brain control's eye movements, or how the neurons of primary visual cortex learn to respond to moving, oriented bars. There are accurate models of how neurons can be stimulated to excitation by injecting currents, or how action potentials, the voltage spikes produced by such excitation, propagate along the axons.
In the area of visual neuroscience alone, an immense amount of theoretical and computational work has been done over the last two or three decades. In fact, there was even an effort to consolidate all that knowledge and host all relevant data on a single online site under the Visiome Project (Usui 2003). A grand computational implementation of the human visual system can be a wonderful tool for both research and teaching. Such a model will trace the journey of light and its corresponding neural signals through the complex pathways of the visual system and unravel what happens to these signals in each visual area.
Starting from the optics of the lens, the model would describe the formation of the retinal image. It would then compute the responses of the photoreceptors – the rods and cones – to the retinal image and calculate the electrical version of that optical image. Then, the neural code of this image in the form of trains of action potentials that flow down the optic nerves would be computed. The significance of the segregation into the parvocellular and magnocellular pathways at this level must be revealed by the model. Further, the processing of the retinal signals in the lateral geniculate nucleus of the thalamus would be described. Then, the preliminary processing steps in the first cortical stopover, the primary visual cortex, would be delineated. Onward, the model should depict how the primary visual information diverges into the ventral and dorsal pathways, and reveal how the objects are recognized at the end of the ventral pathway, while the spatial information is extracted by the second pathway.
Such a model, when it becomes available, can have a profound impact on how an ophthalmologist understands visual processing in the brain. It is highly desirable for such an initiative to be undertaken by experts in India. In combination with the growing arsenal of neurotechnology, with its impressive array of brain stimulation techniques, such a comprehensive computational model may show pointers to novel treatments of visual ailments that have their roots, not only in the obviously visible organ of the eye, but also deeper in the dark and mysterious recesses of the visual brain.