Computer program could detect autism
Before Saiqa Akhter dominated autism headlines this week, science produced news that could assist developing an objective form to detect autism, marking the second time this year I’ve come across an article that discovers new possibilities.
The first brought forth the theory of using urine tests to diagnose autism. This time, scientists say they’ve designed a computer program that can distinguish speech patterns between autistic and normal children, correctly identifying more than 85% of its subjects. Scientists working on the study discovered that autistic children mangled their syllables for much longer than non-autistic children, making a diagnosis easy for the computer program. However, those involved in the study recommend caution: the lead researcher doesn’t believe the program should be the sole measurement in determining who is autistic.
The article may be another sign of an incoming phase in autism coverage. A CNN article a few years back highlighted the primary problem of no objective method of diagnosing autism. A couple years back, ABC News reported a study that found certain genes in an autistic person’s DNA lie dormant. With two discoveries this year, it’s clear there’s a focus at finding something that will make autism diagnoses foolproof (there’s some skepticism at the rapid increase in autism diagnosis rates). While the Internet posts its share of contemporary issues, the medium is also proving its use at publishing articles paving new paths for the autism community, and it goes well beyond Google searches. Home pages of various e-mail and Internet service providers have a page that features major news stories at the moment, much like the front page of local and national newspapers did before the delivery of news rapidly increased speed. Because these sites update their top stories every few hours, topics ignored by other outlets have a place to carry their information.
It would be a safe bet to suggest more objective methods of looking for autism in children will appear, although how coverage will branch after that is unknown. Reporters may choose to focus on how soon objective tests will be put in place, others may weigh pros and cons against the status quo, and there may be a chance of hearing stories about psychologists’ appointments dropping because their services will no longer be required to provide an autism diagnosis. Of course, they will still be needed to help manage autistic people’s behaviors. In any case, expect computers to be at the forefront in finding an indisputable form of detection.
Share this:
Discover more from Autisable
Subscribe to get the latest posts to your email.
Additional diagnostic tools are always good, in any field, but if this in any way damages the ‘growth rate’ of autism (‘there’s some skepticism at the rapid increase in autism diagnosis rates’) it will be opposed by many who have much vested in seeing as many people as possible diagnosed as being on the spectrum. I have my doubts (and I am on that spectrum too) whether the ‘most-normal’ end of AS is really non-neurotypical at all, and not something else of a different order.
The more research, the better. But obviously, humans need to evaluate every case individually – even if the computer has picked out many autistic children correctly. Every person is different.