Autism image used for representational purposes only.
Autism image used for representational purposes only. 

New perspectives on autism

Shrimansi Kaushik

HYDERABAD: This year’s observance of World Autism Day on April 2 highlighted the theme ‘Moving from surviving to thriving’, stressing the importance of shifting focus from mere coping with autism to fostering an environment where autistic individuals can flourish.

Research in Autism Spectrum Disorder (ASD) has surged significantly in the past decades and the latest developments hint at the role of artificial intelligence (AI) in fundamentally changing how we view and support individuals with ASD.

The role of AI and technology

“AI algorithms are enhancing diagnostic accuracy. Computer games, robotics, and virtual reality offer new ways for therapy tailored to individual needs,” says Dr Dedeepya Puskur, Developmental Paediatrician and Clinical Head at Fernandez Child Development Centre.

Citing some recent studies, she specifies how AI is helping in the detection and screening of ASD. “In a 2019 study published in the Journal of Autism and Developmental Disorders, Bussu G and others demonstrated the potential of AI by using a support vector machine (SVM) to predict ASD diagnosis based on developmental evaluations during infancy, enhancing the classification of atypical development.

Furthermore, in another research paper published in Autism Research, Liu and others investigated the use of facial expression software as indicators for ASD classification, using eye-tracking data to discern differences between ASD and non-ASD individuals. Their study achieved an impressive overall accuracy of 88.51%, showcasing the efficacy of AI in diagnosis,” Dr Puskur said.

She also mentioned another study that used AI to differentiate between ASD and attention deficit hyperactivity disorder (ADHD). “Duda and others published their research in Psychiatry in 2019, which extended the application of AI to differentiate between ASD and ADHD using the Social Responsiveness Scale (SRS), highlighting AI’s potential to distinguish between neurodevelopmental disorders,” she added.

These examples illustrate the multifaceted role of AI in revolutionising ASD screening and diagnosis, and identification of individuals on the spectrum. However, Dr Puskur mentions that integrating AI into healthcare systems requires careful consideration of feasibility and accessibility. Concerns about screen time necessitate further research and awareness to mitigate potential risks.

Dr Garima Vegivada

Innovations in speech therapy

While AI has extended the role of technology in diagnosis, some innovations in traditional methods have also helped provide better treatment. According to Dr Garima Vegivada, Clinical Director, Child Development Unit, Hear ‘N’ Say clinic, Gestalt Language acquisition can help children with ASD learn speech and language.

“Gestalt language processing is a way to understand how children with autism acquire language. About 75-85% of autistic children are gestalt processors. Children have echolalic speech which they pick up from songs, movies, books, or from family members around them,” she said.

The word ‘gestalt’ means ‘whole,’ so gestalt units are typically long ‘wholes’: whole sentences, whole songs, whole stories. At first, children produce ‘chunks’ or ‘gestalt forms’, without distinction between individual words. As they understand more about syntax, they can break down these ‘gestalt forms’ and begin to recombine segments and words into spontaneous forms. Eventually, the child can formulate creative, spontaneous utterances for communication purposes.

Speech therapy for gestalt language learners

“The ultimate goal in speech therapy for children who are gestalt language learners is to eventually use self-generated language,” said Dr Vegivada. “It is most effective when language is targeted in natural and engaging contexts, like children’s play,” she added.

Some tips for responding to gestalt language

Respond! Smile, nod, or even just repeat it back to acknowledge that you know it is a communication attempt.

Don’t take the scripted comment seriously (“it’s on fire!” may have a different meaning to the child than something actually being on fire/hot).

Be a detective — Ask yourself, ‘What are they communicating to me when they say this script?’ Once you discover the true meaning, you will be able to target ways to increase their self-generated language through modelling your own language.