AI device helps diagnose autism in children

Geralt / Pixabay

Source: Geralt / Pixabay

A new peer-reviewed study is published in digital medicine npja Nature Portfolio, showing results of a clinical trial in it Artificial intelligence Machine Learning (AI) as a Medical Device (SaMD) software has helped primary care providers assess whether young children have autism spectrum disorder (ASD).

Autism is a disorder of the whole body accompanied by common diseases that include worryAnd the depressionAnd the Attention deficit and hyperactivity disorder (ADHD), schizophrenia, Two-way disorderSleep disorders, digestive disorders, eating and feeding problems, seizures. Autism affects all ethnic groups, and boys are four times more likely to develop autism than girls.

The World Health Organization (WHO) estimates that one in every 100 children globally has autism spectrum disorders. One in 44 eight-year-old American children was identified as having autism in 2018 according to the Autism and Developmental Disabilities Monitoring Network (ADDM) of the US Centers for Disease Control and Prevention (CDC).

This peer-reviewed study evaluated software as a medical device called Cognoa ASD Diagnosis Aid that takes advantage of machine learning for artificial intelligence and consists of a mobile app for caregivers, video analytics portals, and health care providers. Stanford Associate Professor of Pediatrics and Biomedical Data Science Dennis Wall, Ph.D., the scientific founder of Cognoa, created an AI machine-learning algorithm to categorize data from interviewer-based scans that clinicians give a caregiver about autistic symptoms while he was at the Bioinformatics Center at the College of Medicine. Harvard Medicine.

The machine learning algorithm was originally developed in Dr. Wall’s lab. The AI ​​was trained on the Autism Speaks’ Autism Genetic Resource Exchange (AGRE) database and then validated using the same database along with data from the Simons Foundation and the Boston Autism Consortium. Dr. Wall’s AI classifier was 92 percent accurate in predicting those without autism spectrum disorder according to a previous study published in 2012. The AI ​​was improved to include a tool that clinicians use based on direct observations of a child called the Autism Diagnostic Monitoring Schedule (ADOS).

The version of AI evaluated in this current study is in its fourth generation and has been enhanced by subsequent research and development. This study evaluated the ability of an AI-enabled device to help medical professionals diagnose autism spectrum disorder in children aged 18 to 72 months with parent or health care provider concern about developmental delays. The predictions generated by the machine learning device were compared to a human clinical diagnosis based on DSM-5 Standards are validated by one or more reviewing medical professionals.

In 425 study participants, an AI machine learning algorithm produced a diagnosis in 32 percent of patients as either “autism spectrum disorder positive” or “autism spectrum disorder negative.” The predictive accuracy of AI is 98.4% for children with autism, and 78.9% for those without autism spectrum disorder. For the 68 percent of the children where AI gave “indeterminate” results, 91 percent had one or more neurodevelopmental conditions.

To put the “non-specific” classifications in context, it is useful to note that one or more mental health conditions often accompany people with autism. according to Autism and Health: A Special Report by Autism SpeaksEpidemiological studies estimate that 54-70 percent of people with autism have at least one psychiatric condition. For every autism speaking, the most prevalent mental health condition among autism he is Attention Attention deficit hyperactivity disorder (ADHD). The report notes that in various other studies conducted in the past 10 years, an estimated 30-61 percent of people with autism also have ADHD, which is much higher than about 6-7 percent of the general population with ADHD. Motion and attention deficit from the CDC. Among people with autism, an estimated 11-42 percent have an anxiety disorder or more, 7 percent of children and 26 percent of adults have depression, 4 to 35 percent of adults have schizophrenia, and 6 and 27 percent have schizophrenia of bipolar disorder.

Copyright © 2022 Cami Rosso All rights reserved.