How Artificial Intelligence is breaking barriers in Autism Diagnosis and Care
For any parent, the early years are a most valuable countdown of “firsts” of his or her precious child: the first step, the first clear word, the first beautiful smile, and quite a few other firsts as well. Yet for all that, for some families, that joy is overshadowed by a growing, quiet, but disturbing intuition that something is even a little bit different. Perhaps a child is not responding to his or her name, or the little one seems to be more interested in the spinning wheels of a toy than a game of peek-a-boo, or even avoids normal social responses.
In many countries, especially in the developing world, the road from that first “gut feeling” that there is something wrong, to a formal diagnosis of Autism Spectrum Disorder (ASD) is often a long and exhausting journey. While doctors can often identify autism in children as young as 12 to 18 months, the average age of diagnosis in our communities still hovers around four years. In these critical years, when a child’s brain is most like a machine ready to learn and adapt, time is of the essence and is the most valuable resource a family has.
Today, a new “algorithmic dawn” is offering a shortcut to really cut that delay. Artificial Intelligence (AI), the very same smart technology that helps us navigate traffic, suggest a new song, or help people with ChatGPT, is moving out of the lab and into the children’s nursery. By acting as a digital “magnifying glass”, specifically designed AI tools can now spot subtle patterns in a child’s gaze, some little quirks in the rhythm of their babbling, or the way they move, often much faster than the human eye can. Then the machine can issue a warning signal and indicate that further action and a proper evaluation are necessary. This is most certainly not about replacing the brain, the heart and the expertise of a paediatrician; it is about providing “Smart Help” that can be accessed from a smartphone in a family living room. For millions of “little ones on the spectrum”, most notably in the developing world, this technology is turning a journey once defined by waiting, uncertainty and even tears, into one of proactive care and even brighter horizons. The time gained is most certainly a very valuable window of opportunity.
What is the “Spectrum,” and Why Does Time Matter?
Autism is described as a “spectrum” because it affects many children somewhat differently and to varying degrees. Some children may have advanced technical skills but struggle to hold a conversation; others may be non-verbal or have intense sensory sensitivities. It can be very mild or very severe, and perhaps everywhere in between as well.
The common thread is that the brain develops differently in these affected children. This is why Early Intervention is the gold-standard goal. During the toddler years, a child’s brain is incredibly “plastic”, meaning that it is a highly adaptable and ready to learn type of organ. Starting therapy and management strategies during this valuable period of opportunity can fundamentally change a child’s future life path.
The problem, to a certain extent, is that traditional diagnosis of ASD is a slow, manual process. It requires intensively trained experts to watch a child play for hours and fill out complex checklists. In many countries, including Sri Lanka, where there is a massive shortage of these highly qualified specialists, the waiting list for a consultation alone can take months or even years. These doyens are rather thin on the ground and even when available, are heavily overworked.
Enter the AI Revolution: Seeing the Unseen
AI certainly does NOT replace doctors, but it acts like a high-powered magnifying glass. By using “Machine Learning”, computers can analyse massive amounts of data to find tiny patterns that the human eye might miss. Here is how it is changing the game:
1. Tracking Gaze and Smiles
One of the earliest signs of autism is how a child looks at the world. AI “Computer Vision” can analyse a simple video of a child playing. It can track exactly where the child is looking. Does the child look at a person’s eyes when they speak, or are they drawn to the spinning wheels of a toy in the corner? AI can quantify these “social attention” patterns in seconds and add them to a cache of things that ring warning bells.
2. The Sound of a Voice
Did you know that the “music” of a child’s speech can hold clues? AI can listen to the pitch and rhythm (called prosody) of a child’s voice. Children on the spectrum sometimes have a “flat” or monotonic way of speaking. AI algorithms can measure these vocal biomarkers with incredible precision, helping to flag concerns long before a child is old enough for a full conversation.
3. Movement and Play
Repetitive behaviour, like hand-flapping or rocking, are core traits of ASD. Sensors in smartphones or simple video analysis can now categorise these movements objectively. Instead of a parent trying to describe how often a behaviour happens, the application or ‘app’ provides a clear, data-driven report for the doctor.
Innovation at Home: India’s Digital Solutions
The most exciting part of this technology is that it does not require a million-dollar lab. In India, where smartphone use is booming, several “homegrown” apps are bringing specialist-level screening to rural and urban homes alike.
Apps like CogniAble, which give parents a step-by-step intervention plan based on the child’s specific needs, or START, a tablet-based tool used by local health workers in areas like Delhi slums to spot risks via simple games, or LEEZA.APP, which offers free AI screening to remove the “money barrier” that keeps many families from seeking help, or AutismBASICS, which provides thousands of activities and a milestone tracker to help parents manage daily therapy at home, are just a few of the programs in use at present. These tools are “democratising” healthcare. A mother in a remote village with a basic smartphone can now access the same level of screening logic that was once only available in a major city hospital.
Beyond the Diagnosis: A Robot Tutor?
The role of AI does not stop once a diagnosis is made. It is also becoming a tireless “co-therapist.”
For many children with autism, the human world can be unpredictable and overwhelming. AI-powered “Social Robots” or interactive apps provide a safe, predictable environment. These “Robo-Therapists” do not get tired, they do not get frustrated, and they can repeat a social lesson even 100 times until the child feels comfortable.
Furthermore, for children who are nonverbal, AI-powered communication apps serve as a “voice”. These apps use smart technology to predict what a child wants to say, allowing and facilitating them to express their needs and feelings to their parents, even for the very first time.
The Human Element: Proceed with Care
As bright as this dawn is, experts warn that we must move forward carefully and most intelligently.
= Privacy: Because these apps collect sensitive videos and data about children, keeping that information secure is a top priority.
= Cultural Differences: An AI trained on children in the US or Europe might not perfectly understand a child in Sri Lanka. We need “diverse local data” to ensure the algorithms understand our local languages, gestures, and social norms. Many of these programs need to be home-grown or baked at home in Sri Lanka.
= The Human Touch: Most importantly, we need to always remember that AI is a tool, not a replacement. A computer can spot a pattern, but it cannot give a hug, provide emotional support to a struggling parent, or celebrate a breakthrough with the same joy as a human therapist.
A Brighter Future
We are moving toward a world where “waiting and seeing” is no longer, and quite definitely, not the only option for parents. By combining the heart of a parent and the expertise of a doctor with the speed of an algorithm, we can ensure that no child is left behind because of where they live or how much money they have.
The “Algorithmic Dawn” is not just about code and data. It is about giving every child the best possible start in life. It is the main principle on which Hippocrates, the Father of Medicine, all those centuries ago, based all his postulations on how physicians should work.
The “Red Flag” Checklist: 18 to 24 Months
The American Academy of Pediatrics recommends screening all children at 18 and 24 months. If you notice several of these signs, it is time to use an AI screening app or consult your paediatrician.
Communication and Social Cues
= The Name Test: Does your child consistently fail to turn around or look at you when you call his or her name?
= The Pointing Test: By 18 months, most toddlers point at things they want (like a biscuit) or things they find interesting (like a dog). Is your child using your hand as a “tool” to get things instead of pointing?
= The Eye Contact Test: Does your child avoid looking at your face during social interactions or during play or when being fed?
= The Shared Smile: Does your child rarely smile back when you smile at him or her?
Behaviour and Play
= The Toy Test: Does your child play with toys in “unusual” ways? (e.g., instead of rolling a car, they spend 20 minutes just spinning one wheel or lining them up in a perfect, rigid line).
= The Routine Rule: Do they have an extreme “meltdown” over tiny changes, like taking a different route to the park or using a different coloured cup?
= Repetitive Motions: Do you notice frequent hand-flapping, rocking, or spinning in circles, especially when they are excited or upset?
The “Golden Rule” of Regression
Finally, an extremely important rule for concerned parents to follow.
If your little one had words (like “Mama” or “Dada” or “Amma” or “Thaththa” or Thaii/Amma or Appa) or social skills (like waving “Bye-Bye”) and a beautiful social smile etc, and then SUDDENLY STOPS USING THEM, that could be a most significant red flag. In such situations, the standard advice would be: Please consult a doctor immediately.
by Dr B. J. C. Perera
MBBS(Cey), DCH(Cey), DCH(Eng), MD(Paediatrics),
MRCP(UK), FRCP(Edin), FRCP(Lond), FRCPCH(UK),
FSLCPaed, FCCP, Hony. FRCPCH(UK), Hony. FCGP(SL)
Specialist Consultant Paediatrician and Honorary Senior Fellow,
Postgraduate Institute of Medicine, University of Colombo, Sri Lanka.
from The Island https://ift.tt/mWEjoPw



