Images,Imaging,Imagination

Light of the appropriate wavelength is processed through the retina and optic nerve to reach the occipital cortex and thence consciousness. The processing of signal should be straightforward although elusive in its detail. Light passes through the cornea, gets focused through the lens, and strikes the retina, where dim-light rods or bright-light cones act as photoreceptors and synapse in the retina before the careful light map is sent to the visual cortex in the occipital area of the brain. Our species is competent to process light in the spectrum from red at 390 nm to violet at 750 nm. Our view of the world is deeply intertwined in our ability to process this lovely rainbow of the electromagnetic spectrum. From the relatively simple representation of light reflected from objects, we form our sense of our surroundings and process that information into a richly intricate, metaphoric, symbolic logic. An image is an artifact that is similar to a subject and can be perceived by our visual apparatus. It is based on light we can perceive in the visual spectrum. Imaging is a matter of creating images for our eyes to behold and process. Now it gets complicated.

We heavily rely upon all manner of sectors of the full electromagnetic spectrum to acquire an image, even in regions completely alien to our visual receptors. Through imaging we can convert radio waves all the way up to gamma rays into coherent and relevant datasets to inform our visual cortex through signal processing. Radio sources in the universe are captured by radiotelescopes and presented as images for us to contemplate and study. Infrared and ultraviolet are made entirely visible by processing. X-rays are made visible on traditional photographic chemical plates or photonic processors to provide us with detailed information of structure based upon the relative penetration of the radiation through a material such as bone or air. We can excite material through their hydrogen bonds in a strong magnetic field to resonate and emit energy that can be recorded, processed, and give us an image of structure in a coherent visual range. We can induce the collision of matter and antimatter to produce metabolic images through positron emission tomography scans. Positron emitting substrates are introduced into the body, which cause gamma decay in their collisions with electrons, and the gamma radiation is captured, processed, and rendered into a lovely picture. We can introduce many gamma emitters into the body to obtain nuclear scans in a similar mode. In this way we seem to see a specter of an individual in its metabolic vibrancy without regard to overlying or underlying structure. Further, we can make ultrasound energy visible in its reflection from a piezoelectric transducer as it bounces back from body structures to a receiver. This can even be made colorful and indicative of the direction of flow by processing the signal based upon the Doppler effect. We can image the microscopic, molecular, and subatomic by microscopy and produce an image for our vision that is colorful and readily processed by the brain that was prepared only for sunrise to sunset on an African plain serving our bipedal ancestors to hunt and gather.

Ultrasound, nuclear magnetic resonance, gamma radiation, X-rays, infrared, ultraviolet, and radio emissions cannot be seen at all. However, to use our analog brains to understand structure and function all are rendered into some aspect of the visual spectrum so we can see the unseen. Our reference to the visible spectrum for understanding the dimensions, shape, texture, subtlety, and nature of objects in our physical world is unshakable. There is no problem of course with such reliance, but a few observations may be of interest.

First we should consider that the image being prepared for our eye as a dataset will be processed into something we will understand. There is no reason why that image must be a form. Edward Tufte at Yale has made a highly constructive career with works such as “The Visual Display of Quantitative Information.” He has taught us that abstract data can be made palatable and readily understood by the eye if presented with proper graphics and shapes or lines being substituted for raw data. If our visual brain can conceptualize invisible objects, there is no reason not to utilize that part of our perception fully to conceptualize very abstract datasets. Certainly it is common that we can visualize very abstract mathematical, esthetic, and even spiritual matters in our brains. Thus, we are able to manipulate quantum mechanics, the texture of music, the nuances of art, or the feel of religious experiences. However, we are not all mathematically gifted. Some cannot really understand the difference between a sonata and gigue. And many rely heavily upon icons to get the sense of spiritual ideas. What we do most readily is to process visual images made for us and then work from the visual analogs. If we consider the image as a dataset subject, the matter gets a little easier. Let us see F sharp. Let me hear blue. Let us smell darkness. If we think of the image as a dataset, the opportunity to utilize a collage of senses becomes feasible.

We know intuitively that images are artifacts. Therefore, what is brought to us through the eye may be subject to error as well as interpretation. Eye witnesses have been to subject of error certainly from the time of the infamous trial of Sacco and Vanzetti in 1921. We process images internally in a way that matches expectations. In surgery, egregious errors in laparoscopy were found to come from surgeons seeing what they thought they should see through a videoscope, ignoring the real image presented. Further, digital editing can change an image in ways that are deliberately deceptive to misinform. One can never assume that the preparer of an image for our consumption is a neutral party or invariably committed to accuracy. The goal may be to misrepresent for an ulterior motive. Make the image prettier to sell, uglier to induce abhorrence, more colorful to induce acceptance, etc. The error in the image may be the result of simply ignoring the rules of sensitivity and specify. The image may suffer from noise-to-signal ratio problems. The image is indeed not really the object but a dataset representing the object. There are wonderful standards for transmitting images and for image capture and creation. However, the standards are not completely pervasive, and consideration of image as dataset might prepare us to generate better standards for accuracy and fidelity.

We are also accustomed to processing the information we receive through the filter of experience to categorize the image, to internally define the relevance to other images we know, and to anticipate the impact of what we are seeing. Processing the dataset is customary in the huge digital files we get with computed tomography or magnetic resonance images. We are comfortable plowing through the millions of pixels for magnification, slicing at another angle, and enhancing in a hundred ways through the programmed tolerances of the dataset. In managing raw data we think of data mining and search engines to manipulate vast amounts of captured bits to generate a mental picture of the pixels and voxels to permit an understanding. Using the visual cortex for nonvisual datasets may be more natural. However, we are on a poor footing if we rely upon internal processing because of the difficulty of internal error. If we are to visualize datasets as an extrapolation of image recall, most of us will need to be taught and disciplined as to rules to avoid error.

Here is the last point. We can consider an image as not just an object but as a dataset for anything the mind can contemplate. It is imperative that the image be accurate and that its data parameters and limitations are understood and quantified. Our ability to deal with abstract concepts as images is probably quite extensive and grossly underutilized. Now, the biggest hazard to truth is probably not the quality of the image we receive. Rather, the caveat goes to the actions we take as a consequence of the image. When one anticipates the outcomes based upon perceptions, we say an individual has vision. That is generally regarded of course as a virtue. However, the distinction between vision and hallucination can be very subtle!! In our field of telemedicine with its absolute reliance upon transmitted, faithful representation of a distant event, we are at the beginning of understanding the ultimate human–machine interface for information. Imaging is not really different from any other data transfer. Our task is to avoid data hallucination, false processing, and keep to the truth.