So you might be asking yourself "How and why did they do this"?

Back around September 2000, Dr. Carl Cotman asked for my advice on a project which was being undertaken at the Institute for Brain Aging and Dementia at the University of California, Irvine. The ultimate goal of the project was to produce a holographic representation of the human limbic system for the Image and Meaning conference at MIT in June 2001.

Previously, Dr. Cotman's daughter, Cheryl Cotman, had been working, along with neuroanatomist Dr. Jay Angevine for several months on a series of realistic sketches of the limbic system. Sketches of the limbic system are not rare, in fact, you can find one in almost any anatomical textbook produced for introductory psychology to advanced neuroanatomy. However, these sketches were special in that they showed a level of detail and relationship between the various limbic structures which had rarely been seen before. To see an example of one of these sketches click here.

The problem of course is that to produce a hologram, 100's if not thousands of these sketches, from various viewpoints, all perfectly scaled, would have to be made and digitized; this would more than likely take a lifetime to accomplish. So a more efficient method of producing a medically accurate and artistically attractive representation of the limbic system needed to be developed. Cheryl and Dr. Angevine contacted a friend of their's who was proficient with 3-D studio Max (3-D drawing and rendering software) and he tried to draw the limbic system directly within that software. The results were promising, but they resembled more one of the early sketches of the limbic system which were intended to show only the basic shape and approximate relationship between the limbic structures, and not really appropriate for this project. The next avenue of investigation were MRI (magnetic resonance imaging) scans. MRI is used for a number of purposes from medical diagnosis to scientific investigation. Unfortunately, the level of resolution provided by MRI is still not quite accute enough to reveal all the brain structures of interest, and still be able to exclude the structures which were not of interest. One of the best examples of MRI attempts to "tease out" limbic system structures can be found here at UCLA's Laboratory of Neural Imaging. This is an awesome resource for anyone interested in neural imaging with MRI.

At this point, Cheryl, Dr. Angevine and Dr. Cotman were starting to run out of ideas on how to attack the problem, and out of time. This is the point where I became involved. It was really a chance conversation that I had with Dr. Cotman; he had called our home to talk to my wife who works with him at the Institue of Brain Aging and Dementia. We talked about my interest in computers and whether I was interested in having a look at what they had done, and if I had any suggestions. At the time, I was not working (due to INS restrictions which is a whole other story), so I relished the opportunity to do something constructive with my time. I went in and met with Dr. Cotman showed me what had been done so far and we discussed the project and threw around a few ideas. I agreed to investigate possible techniques for accomplishing our goal. Little did I know how involved the project would become, and how deep into it I would get.

I went home and started searching on the net, what for I didn't quite know, but I started with an exploration of what had been done by others and if we could simply "borrow" their work to produce our hologram. It quickly became evident that nobody had done this type of work with the human limbic system or any brain system for that matter. Although there were plenty of MRI sites none had produced images sufficient to produce the work we wanted; we were in uncharted territory.

Next I started to investigate software packages which could be used to produce 3-D images of biological entities. Luckily, one of the first packages I found looked capable of doing the job. The package is called "Surfdriver". This package is capable of taking digitized serial slices of tissue and combining them into a 3-D object. Normally Surfdriver would be used to illustrate the 3-D structure of a blood vessel or a legbone etc. the use I had in mind would certainly push surfdriver to it's limits. I persuaded Dr. Cotman to purchase Surfdriver along with a copy of the "Visual Human Project". Unfortunately after "messing around" with Surfdriver and the Visual Human project I realized that teasing out the limbic system would be next to impossible because many of the limbic structures could not be seen. The Visual Human method of tissue preparation did not include any kind of staining. This is not a criticism of the Visual Human project as it is absolutely amazing, it's just that it didn't suit the needs our project. Luckily Dr. Angevine saved the day as he had a beautiful series of stained serial sections of the human brain from the Yakovlev collection, all convieniently photgraphed on 35mm slides. We couldn't have hoped for a better resource. I scanned them into a computer readable format using a high resolution slide scanner at the Institute for Brain Aging and Dementia.

We had a teleconference meeting; Dr. Cotman, Cheryl, Dr. Angevine and myself and decided on a course of action. Dr. Cotman lent me a Wacom drawing tablet, and I, taking neuroanatomical direction from Dr. Angevine, was to trace the limbic structures on the 200 or so slides and put together the limbic system in 3-D using Surfdriver. When that was done, Cheryl was going to add color and provide some artistic direction. This of course was easier said than done! I spent about a month or so using various resources and many phone calls to Dr. Angevine in Arizona trying to accurately represent the limbic system, but it just didn't look quite right. I have a background in neuroscience, so am no stranger to neuroanatomy, but to create an accurate 3-D model of the limbic system from 2-D photgraphs is not as easy as it sounds. We arranged another meeting, one which brought Dr. Angevine, Cheryl and myself together and we did a little brainstorming. It was decided that we would print out the brain section slides, and Dr. Angevine would manually trace each section on tracing paper to outline each limbic structure. He would then send them to me and I would use his tracings as the resource to get all this into the computer using Surfdriver.

Once provided with Dr. Angevine's tracings, the project moved along quite swiftly. Within about a month, I had a reasonable 3-D model of the limbic system including the cortex. Surfdriver is a great program, however, it does have it's limits. It's rendering and editing ability is not up to par with some of the other commercial 3-D rendering programs. It's main purpose, as I stated before, is to render gross structure rather than exquisite detail, fortunately the makers of Surfdriver forsaw the need to do more with 3-D models than what they offered in their program, and provide a facility to export Surfdrivers output into other formats. I exported my files in .dxf (autocad) format and then imported them into 3-D Studio Max. A number of things happened in the conversion process, the most disturbing of which was that edges became less rounded making the model look "blocky". However, using some of the tools in 3-D Studio Max, I was able to smooth out the model and remove some of the errors that were made by both the act of manually tracing and those introduced by the file conversion process. We now had an almost perfect 3-D model of the limbic system.

We arranged another meeting between Dr. Angevine, Cheryl and myself to go over some of the final details. There were a few points that Dr. Angevine was not happy with but overall, we were all pleased by the progress. The human brain is not completely symetrical, although it appears so at first glance, so for artistic reasons, it was decided to use only half the model and to make a mirror image of the "best half". After making the needed adjustments, the ball was in Cheryl's court to provide color and "material" to the model.

In the meantime, we had to shop for a hologram company. The first company that we had talked to way back in September could no longer fit us in their schedule as it was already April. We settled on a company called "Holographics North". I can't say enough good things about them. John Perry at Holographics North said it was short notice, but he could accomodate us. In order to produce a hologram from our model, I had to lighten it up with 3-D studio's lighting, and render 180 images of each color band in greyscale. As well, I had to put a calibration marker on each rendering so they could be lined up during the hologram process. Once Cheryl had finished with coloring, I readied the model for the hologram. We had all decided we wanted an animated hologram so as you walk around it, it "moves" from a frontal view to a rear view. Working with John Perry, I created the180 views for each color channel and sent them to him. The hologram was produced and sent directly to MIT just in time for the conference. During the time the hologram was being produced Cheryl worked on a "whiteboard" presentation with one of the original human brain slices encased in glass, which was presented along with the html presentation (seen on this site). As well, I produced a short animation of a full rotation of our model, and a VRML demonstration to go with the html presentation. To see a photograph of the hologram click here.

After the presentation at MIT, I was asked by Dr. Cotman to expand on the animation I made because he wanted to present it at UCI along with the hologram and the other work Cheryl had done. He wanted me to use the animation to convey the theoretical spread the limbic and cortical components of Alzheimer's disease and to add some sound. To do this I went back and edited the model and used lighting to convey the spread of disease. The effect is interesting and raised a few eyebrows when it was presented at UCI in December 2001. The accompanying music is a mixture of Tangerine Dream's theme from the movie Risky Business, and some music made from actual electrical activity in the hippocampus which Cheryl worked on, it's and interesting mix.

There were a lot more steps involved in this project than I have mentioned here, but I have been omitted them to make this a short story :-) The following link is a brief powerpoint-like presentation which gives a summary of the steps we followed if anyone reading this is thinking about embarking on a similar project click here. For a list of the computer and software specifications and a more detailed description of the steps involved click here.