A new study published in Science Translational Medicine describes the work of researchers at San Diego's Rady Children's Institute for Genomic Medicine to rapidly diagnose genetic diseases using whole-genome sequencing.
The new method is powered by automated machine learning and clinical natural language processing and is able to deliver genetic test results to physicians in neonatal and pediatric intensive care units significantly faster than manual and other genome sequencing methods.
Rady's rapid Whole Genome Sequencing (rWGS) system uses a single blood sample to screen a child's entire genetic makeup for anomalies. The technology is integrated with, among several platforms, a program from London-based data analytics firm Clinithink that simultaneously searches a patient's EHR to extract relevant phenotype information, and MOON software from Belgian genomics company Diploid, which uses AI to detect the disease-causing mutation within minutes.
In the study, the rWGS system retrospectively diagnosed rare genetic diseases in 95 children with 99 percent precision when compared to the diagnoses of expert manual interpretation. The system took a median time of just over 20 hours for each diagnosis.
More articles about AI:
Elon Musk says an update on his neurotech company is 'coming soon'
UC San Diego unveils AI-powered tool to monitor runners' heart rates, suggest optimal workouts
FDA approves AI-enabled wearable device for at-home vital sign tracking