By Matthew Phelan Senior Science Reporter For Dailymail.Com
Researchers have developed a method for diagnosing autism which could spare families years of uncertainty and spur crucial earlier treatments.
The new AI analysis can identify the genetic markers of autism via biological activity in the brain, they report, with 89 to 95 percent accuracy.
This new method starts out with standard brain-mapping via magnetic resonance imaging (MRI) before re-analyzing those scans via AI to detect the movements of proteins, nutrients and other processes within the brain that may indicate autism.
‘Autism is traditionally diagnosed behaviorally,’ via a person’s speech, for example, as the medical team behind the process noted. ‘But [it] has a strong genetic basis.’
Today, autism affects one in 36 children, according to the CDC, meaning that more than 90,000 children are born annually with this developmental disorder in the US.
But, autism is notoriously hard to spot and the vast majority of children with the condition won’t be diagnosed until the age of five and show clear behavioral signs.
Worse, that identification process usually involves years of uncertainty, dozens of trips to the hospital and a battery of tests, including speech and language exams, observational interviews and more, which can be stressful for children and families.
The new diagnostic technique, the researchers hope, will soon allow doctors to locate more specific genes responsible for autism — by first revealing the actual biological pathways through which autism changes how a brain grows and operates.
As a spokesperson for a university behind the new method put it, the method ‘cracks the autism code,’ although there is no word yet on when it may enter common use.
Dr Shinjini Kundu, an assistant professor of radiology at Washington University in St. Louis, developed this new machine-learning AI, mathematical brain-modelling technique while she was a grad student-researcher.
The method, named ‘transport-based morphometry’ after the transport of biological matter in the brain, focuses in on identifying patterns linked to key bits of genetic code.
Those sequences of genetic code, called ‘copy number variations’ (CNVs) reveal segments of DNA that have been deleted or duplicated — alterations which have been linked to autism in past research.
‘Some copy number variations [CNVs] are known to be associated with autism,’ according to biomedical engineering professor Dr Gustavo Rohde, who taught Dr Kundu during her PhD studies.
‘But their link to brain morphology — in other words, how different types of brain tissues such as gray or white matter, are arranged in our brain — is not well known,’ according to Dr Rohde, who now teaches at the University of Virginia.
‘Finding out how CNV relates to brain tissue morphology,’ he explained, ‘is an important first step in understanding autism’s biological basis.’
Drs Kundu and Rohde, and their collaborators from the department of neurology at the University of California in San Francisco, published their results developing this new autism identification method in June with the journal Science Advances.
Participants in the nonprofit Simons Variation in Individuals Project, a cohort of subjects with known autism-linked genetic variations, contributed key data used in the new study.
The researchers recruited their ‘control subject’ patients from other medical or clinical settings based on their similarities to the Simons group (such as same age, sex, and non-verbal IQ), to reduce variables that could muddy their results.
Most machine learning methods that plow through medical data like MRI scans, according to Rohde, do not incorporate a mathematical model for the many biological processes that happen to be contained hidden in that data.
Instead, past AI models, were only seeking out patterns to identify abnormalities or statistical anomalies in various patients’ health data.
Dr Kundu’s ‘transport-based morphometry,’ however, could help researchers distinguish even more tell-tale biological variations within brain structures — beyond the deletions or duplications associated with CNVs and autism.
Given reports that 90 percent of all medical data comes from similar imaging, the team hopes this method could help draw new helpful information out of old tools.
‘Major discoveries from such vast amounts of data may lie ahead if we utilize more appropriate mathematical models to extract such information,’ Dr Rohde opined.
‘We hope that the findings,’ he added, ‘could point to brain regions and eventually mechanisms that can be leveraged for therapies.’
Discussion about this post