Skip to navigation Skip to main content Skip to footer

Approved research

Machine learning towards early prediction and diagnosis of Autism Spectrum Disorder (ASD) and Cognitive Function

Principal Investigator: Dr Ashis Kumer Biswas
Approved Research ID: 44022
Approval date: October 28th 2019

Lay summary

This research seeks to use genetic and outcomes information from the UK Health Services and the UK Biobank to reduce the severe implications of Autism Spectrum Disorder cases. Autism is, in part, caused by presently unknown genes. New advanced computer science techniques including Deep Learning (like the techniques that allowed computers to beat world champion chess and Go players), combined with huge new databases like the UK Biobank, have allowed progress in the past ten years in identifying the genes responsible for disorders and autism. This knowledge can help do an early diagnosis of children that may have an Autism Spectrum Disorder. Genetic information will increasingly help alleviate the difficulties that autism patients have. Thus, the research helps to reduce the need for spending on Autism Spectrum Disorders, estimated at more than $268 Billion per year in the USA, and to reduce human difficulties associated with the disorder.