Autistic-like monkeys have inspired local neuroscientists to develop a new machine-learning approach that improves the imaging-based diagnostic accuracy of autism and offers a fresh hope for early diagnosis.
Using genetically engineered macaque monkeys carrying extra copies of the MECP2 gene, researchers used magnetic resonance imaging to scan the brain of transgenic and wild-type monkeys and identified which parts of their brains are affected by MECP2 overexpression. These brain scans provide new insights to researchers trying to identify the core affected areas of the brain in human patients.
As such, local researcher Dr. Wang Zheng and his team from the Center for Excellence in Brain Science and Intelligence Technology in the Chinese Academy of Sciences designed a novel cross-species machine-learning model, which significantly increased the MRI-based diagnostic accuracy of autism to 82.14 percent.
These findings were published in the American Journal of Psychiatry on Wednesday.
An industrial report by one autism research center, Wucailu, shows one in 100 Chinese children is autistic.
Autism remains a poorly understood brain disorder, marked by impaired communication skills and repetitive patterns of behavior.
It often comes with various comorbidities in patients, including obsessive-compulsive disorder.