Introduction: An AI autism diagnosis method has led to a groundbreaking discovery in medicine. American scientists have developed an AI-based approach to diagnose autism with remarkable precision, potentially revolutionizing early detection and treatment. This method offers families the potential to avoid years of uncertainty and enables earlier intervention. The new AI-driven analysis boasts an accuracy rate of 89 to 95 percent in identifying genetic markers of autism through brain activity.

The Goal of the Research

The process starts with a standard MRI brain scan. AI then reanalyzes this scan to detect protein movement, nutrient flow, and other brain processes that may signal autism. Traditionally, doctors diagnose autism based on behavior, but it has a strong genetic foundation. The medical team from Washington University in St. Louis, who developed this AI autism diagnosis method, explained this. Autism currently affects one in 36 children in the U.S., with more than 90,000 children born with the condition each year. Despite its prevalence, autism is often challenging to diagnose. Most children do not receive a diagnosis until the age of five when visible behavioral symptoms emerge. The current diagnostic process is lengthy and stressful, involving multiple hospital visits and numerous tests, such as speech and language assessments.

Scientists aim for this new diagnostic method to help doctors identify the specific genes responsible for autism. It does this by uncovering the biological pathways through which autism affects brain growth and function. For more on the genetic aspects of autism, check out this study on autism and genetics.

The Technique Behind the Discovery

Dr. Shinjini Kundu, an assistant professor of radiology at Washington University in St. Louis, developed this AI autism diagnosis technique. The method, called “transport-based morphometry,” uses AI and brain mathematical modeling to identify patterns. These patterns are associated with genetic code sequences known as “copy number variations.” These variations, which involve deleted or duplicated DNA segments, have been linked to autism in previous research. Learn more about AI’s role in modern medicine.

A Path to Major Discoveries

Some copy number variations are known to be associated with autism, but their relationship with brain morphology has not been well understood. Dr. Gustavo Rohde, professor of biomedical engineering and Dr. Kundu’s mentor, emphasized the importance of understanding this relationship. This understanding is crucial for unraveling the biological basis of autism.

Dr. Rohde noted that earlier AI models lacked the mathematical modeling needed to interpret complex biological processes. However, this new AI autism diagnosis approach could unlock significant insights from existing medical data. “Major discoveries may be ahead of us if we use more appropriate mathematical models to extract such information,” he said. Dr. Rohde expressed hope that these findings could eventually lead to targeted therapies. The Washington University spokesperson praised the method, stating that it “cracks the code of autism.” However, it is unclear when it will become widely available. The research was published in Science Advances. For more on this groundbreaking research, visit Science Advances.