How do we acquire an overall impression of our environment? How does the brain extract different features from sensory sources and then puts them together into a complete form? Computers have major problems in recognizing faces but for the majority of human individuals this is not a challenging task. Most people agree that this attribute of the brain stems from its capacity for parallel processing. Still, getting the information together is far from being a trivial matter. Furthermore, this gathering of information may run amiss in individuals with autism. Such a problem may help explain why autistic individuals have difficulties with face recognition, socialization, or putting objects/animals/things into categories. Temple Grandin in her book “Thinking in Pictures” relates her problem in trying to conceptualize why different breeds of dogs are still classified as canines!
Roberto Llinas once tried to simplify this problem by giving an analogy using crickets (from his book “I of the Vortex”). Imagine a cricket looking for a soul mate. It does so by chirping its love song. Hopefully another far away cricket will hear the same, come together, and mate. However, the scenario changes when the original lonely cricket is joined by a thousand others. If they now chirp at their own whim another cricket far away may not be able to make sense out of the love song. It is only by chirping together that we are able to listen to a coherent signal. The same thing happens with the brain wherein processing of signals appears to be a neatly choreographed ballet. Different areas of the brain need to work together, in parallel, to make sense out of sensory signals.
Imagine again that you are standing in front of a person and trying to recognize who she is. Her face is a composite of different features. In order to recognize her, the brain has to put information together about the color of her eyes and hair (occipital lobe), the symmetry of her face (right parietal lobe), and emotional expression (amygdala). All of these features have to be processed in parallel in order to expedite the process. The brain does this by using brainwave frequencies to flag down each of these areas. Furthermore, the frequencies used for this purpose tend to be relatively high, i.e., >30 Hz or gamma frequencies. So, if different areas of the brain are coupled at the same high frequencies, they are working together in putting different aspect of an experience into a whole.
Recognizing a human face is a difficult endeavor based on putting together a large number of anatomical puzzle pieces.
The previously mentioned flagging down of different parts of the brain at the same high frequencies appears to be abnormal in autistic individuals. Every single study thus reported on gamma frequencies in autism has been abnormal. Looking at the face , with so much information in it, may be akin to looking at the sun. You can only look at the bright sun by squinting your eyes and taking in a small sample of the overall image. In the case of autistics, the complex features of the face are difficult to piece together. Thus, they may remember a person based on individuals aspects of his/her features, e.g., the type of glasses they are wearing or the shape of their lips.
In autism, a reduction in the total number of long connections and the resultant re-shuttling of information through pathways of increased synaptic length limits the number of neurons in disparate brain regions that are capable of following the faster rhythms, e.g., gamma frequencies. Furthermore the emphasis in shorter connections and lack of inhibitory surround for minicolumns allow for local rhythms to invade adjacent territories. Most of this reasoning was reviewed in earlier posts, see https://corticalchauvinism.wordpress.com/2013/02/06/symptoms-of-autism-part2-how-we-think-depends-on-how-our-brains-are-wired/
The relevance of findings many times depends on their explanatory as well as predictive powers. That is why after publishing our initial findings on minicolumnar abnormalities (Neurology 58(3):428-32, 2002) I decided to write another article predicting previously unexplored phenomena that would develop from our described pathology (Casanova et al., 2002). In that particular article I made mention that minicolumnar pathology would itself create abnormalities of cortical connectivity (white matter) as well as gamma frequency abnormalities. Given our previous posts, it should come as no surprise to the readers that Transcranial Magnetic Stimulation (TMS) serves to correct gamma frequency abnormalities, see https://corticalchauvinism.wordpress.com/2013/01/27/why-use-transcranial-magnetic-stimulation-tms-in-autism/
References
Casanova MF, Buxhoeveden DP, Brown C. Clinical and macroscopic correlates of minicolumnar pathology in autism. J Child Neurol 17(9):692-5, 2002.