Hi,
I saw about this competition which demands ideas regarding developing technologies that help kids with autism communicate. I gave an Idea and so, I present it here.
Idea:
In autistic kids, certain regions of brain are over working which results in their over-active nature and fascination for patterns.
In neurofeedback , what is done is that their brain activity is displayed to the subject seconds later they take place and the subject tries to consiously reduce the activity of that part. This sort of neurofeedback methods have been shown to work in the case of autistic patients. But the subjects need to be old enough to comprehend what the display means and what change is required. Also, in many cases, there is learning inability associated with autism which increases the difficulty of the above method even more.
A typical characteristic of autistic kids is their peculiar nature of social communication and obsessional difficulties. If a tool can be provided which can help them in communicating, this would greatly increase their integration in the society. Also, this tool must be easy to use and relevent even in small kids.
The diagonisis of autism can be made as early as 18 months but the therapy with neurofeedback can be given only when the kid is atleast 4.5 years old (and even later if the kid has learning difficulties). So, here I might try to suggest a method of communication which can be used as early as possible and which is based on neurofeedback methodology.
First the activity of brain is recorded using some attractive caps to which the kid does not hesitate (for this, give the kid time to get habituated to).
After that, the activity of the brain is recorded . Now, it has to be displayed in some form which the child is - 1.) Attracted to and 2.) knows which shape/state of the display is more favorable and which is less. For this, I would like to suggest an example of display .. As the autistic kids have a fascination for patterns and aligning colors and shiny objects, we can do like this .. whenever, the activity of the brain is reduced (As is the goal of the neurofeedback), we can display a pattern which autistic kids prefer. And we display change and unfavorable patterns whenever the kid is not trying and the activity of brain is not as desired. This way, we use the natural preference of autistic kids to describe the goal to them in neurofeedback.
Eventually, as the kid grows, he/she might develop liking towards something and that object's image, etc. can be used in the display.
Now, the goal is to use this for using this for social interactions.
One of the problems of autistic kids is that they do not hear and are unresponsive to names, etc. So, weather a child is listening or not can be determined by studying the activity in brain or by the behaviour . Either way, this can be reflected in the display described above combined with the level of brain activity.
Another problem is the problem that they do not recognise humans and do not make eye contacts (or quick strange eye gazes). This problem can also be targetted in a way similar to the above method..
Sometimes, they get irritated by the changes.. So, during the changes, the pattern displayed must be unusually (like the most prefereable ever) so that the change is compensated as much as possible. I doubt that this will work or not, and also, if it works, if it is advisable for learning or will it make the kid more and more dependent on the machine rather than improving his/her condition.
Sometimes, there is preference for some voices over others.So,there must be a feature to speak using other's voices or in voices similar to a known voice to the kid. This can be a great help for a complete stranger in communicating.
Autistic kids are known to merely imitate sounds in language rather than form their own sentences. The same holds for someone else's actions also. I am not sure if the kid might just merely replicate earlier behaviour to get desired display. This might be a problem if the display is only 'behaviour' based. So, taking brain activity into account might be useful for this case.
.......................................................................................
One can also find the idea at the competition web page (and comment there also) :
Autism Connects