Autism research breakthrough: EEG signals in infants can accurately predict autism from 3 months old

Findings pave way for critical early intervention and treatment

SAN FRANCISCO (May 1, 2018) -- A new study in the journal Scientific Reports shows that nonlinear analysis of EEGs, an inexpensive and noninvasive mode of measuring brain electrical activity, can not only predict or rule out autism spectrum disorder (ASD) in infants as young as three months old, but also predict the severity of the outcome. Current standard diagnoses are not reliable before two years of age and are often not done until 4 years or later.

Led by University of San Francisco (USF) associate professor of Health Informatics and Clinical Psychology William Bosl, PhD, together with colleagues at Harvard Medical School and Boston Children’s Hospital, where Dr. Bosl holds visiting faculty appointments, the study predicted occurrences of ASD in infants with greater than 95% accuracy by analyzing EEG (electroencephalogram) signals with nonlinear physics and machine learning algorithms. ASD diagnoses are currently made solely based on behavioral symptoms.

With just released findings from the Centers for Disease Control and Prevention stating that autism prevalence has increased by 15% in the U.S. in the past two years to 1 in 59 children, the importance of earlier diagnosis is key for successful early intervention and healthy brain development.

“Detecting the emerging disorder before a child begins to show symptoms of autism is key,” said Bosl, director of USF’s Health Informatics program and visiting associate professor of pediatrics at Harvard Medical School. “Brain development precedes the emergence of behavioral characteristics. Measuring atypical changes in the brain could open a new window to early intervention that might enable prevention. That may be more successful than trying to reverse symptoms that have already emerged after the brain has developed in an atypical way.”

Bosl’s study found that EEG measurement is a promising technology for monitoring neural development in a broad population of children. The research was conducted at Boston Children’s Hospital, where research participants were recruited.  EEG sensors arrayed in a safe cap-like net were placed over a child’s head to record brain activity. Breakthrough computer algorithms developed by Bosl analyzed minute patterns in the brain activity for clues. EEG technology is relatively inexpensive, which makes it affordable for health clinics and therefore lowers barriers to testing access. It also requires only a few minutes of a technician’s time to complete an analysis. Bosl and his team hope that EEGs can become widely used for routine pediatric “brain checkups”.

“Brain developments precede the emergence of observable behavioral symptoms, perhaps by months or years,” states Bosl. “Because of that, changing of brain patterns should precede later emergence of autism-related behaviors.” The next steps, explains Dr. Bosl, are to test the methods in routine pediatric well-baby checkups. “That will require significant funding for definitive clinical trials. But the result of this work could have significant impact on our treatment of autism.”

The paper was co-authored by Charles Nelson, PhD, director of the Laboratories of Cognitive Neuroscience at Boston Children’s Hospital and Richard David Scott Chair in Pediatric Developmental Medicine Research at Harvard Medical School, and Helen Tager-Flusberg, PhD, Director of the Center for Autism Research Excellence at Boston University.

The study was supported by National Institute of Mental Health (R21 MH 093753), the National Institute on Deafness and Other Communication Disorders (R21 DC08647 R01 DC 10290) and the Simons Foundation.

Journalists interested in speaking with study author William Bosl, PhD should contact Kellie Samson at (415) 422-2697 or

About the University of San Francisco

The University of San Francisco is located in the heart of one of the world’s most innovative and diverse cities and is home to a vibrant academic community of students and faculty who achieve excellence in their fields. Its diverse student body enjoys direct access to faculty, small classes, and outstanding opportunities in the city itself. USF is San Francisco’s first university, and its Jesuit Catholic mission helps ignite a student’s passion for social justice and a desire to “Change the World From Here.” For more information, visit

About the Master of Health Informatics Program

Health Informatics at the University of San Francisco is a rigorous program that attracts an international student population from diverse healthcare and technical backgrounds. It is designed to immerse graduate students in data science, biostatistics, and biosignal processing, while grounding them in clinical and consumer health informatics and project leadership skills. Graduates have gone on to challenging positions as health data analysts with major medical centers such Kaiser, UCSF Medical Center, and Genetech. Other students have continued to pursue educations in PhD or MD/PhD programs. For more information, contact Tamara White: