Two Whitworth University professors say they're developing software they hope will help improve the diagnosis and treatment of patients with autism-related disorders.

The professors, Susan Mabry and Betty Fry Williams, have teamed up to combine the knowledge of autism Fry Williams has learned as a special-education educator and Mabry's research in computerized medical information to develop a three-part software suite to diagnose and track treatment of what is known as autism spectrum disorder.

The term autism spectrum disorder, or ASD, refers to a variety of complex neurological disorders marked by varying levels of difficulty in communicating and relating socially to others. Frequently, they're are accompanied by other developmental disorders. The National Institute of Mental Health says conditions covered by the autistic disease spectrum include syndromes that mimic autism but aren't autism, pervasive development disorders, and others.

The Whitworth professors' research so far has been conducted with the help of students and has been funded by the university. Fry Williams is the current Edward B. Lindaman Chair appointee, an endowed, four-year position at Whitworth that supports her research efforts, and both Fry Williams and Mabry plan to seek additional money to continue development of their diagnosis and treatment system.

A key motivator in the research, they say, is that the incidence of autism spectrum disorders is rising so much that the U.S. Centers for Disease Control has deemed it a public health risk. In a recent report, the CDC said that one in 100 8-year-olds in the U.S. has been diagnosed with an autism spectrum disorder, compared with one in 150 such youths just two years ago.

Fry Williams says it isn't clear why the number of diagnoses is increasing, although researchers are looking to a possible combination of genetic predisposition and triggering environmental factors as a way to explain the rising incidence.

Because the condition has become so much more prevalent, the American Association of Pediatrics now recommends that pediatricians screen every child under 3 years old for autism spectrum disorder.

"We know if we can identify them as young as possible and intervene early, we can make a person as capable as possible as an adult," Fry Williams says.

Preliminary screening involves evaluating a number of possible symptoms within three different cognitive and developmental areas, although autism spectrum disorder often isn't diagnosed before the age of 3, she says.

One of the basic issues Fry Williams has sought to address with research in autism is that diagnosis is difficult and that many pediatricians don't have the tools or expertise they need to do that analysis in the brief time they normally see a child for a checkup.

"Most pediatricians aren't trained to diagnose what is a very complex issue," Fry Williams says.

She and Mabry have completed prototypes of two components of a three-piece software suite that they hope will aid in preliminary screening and tracking of treatments for patients with autism spectrum disorder. They are seeking funding to continue developing the third, and most complex piece: an in-depth diagnosis tool for psychologists and other professionals to formally diagnose a range of symptoms of autism spectrum disorder.

The software uses two forms of artificial intelligence, as well as probability, to collect and weigh data and help create a more in-depth diagnosis. They hope it also will help evaluate treatment options in use and their overall effectiveness, based on data provided by physicians and parents of children already being treated for autism spectrum disorder.

"I think this is a system that will fill a void," Fry Williams says.

Fry Williams first became interested in autism spectrum disorder about 15 years ago, when she noticed that a lot of the children in a preschool special-education program she was running exhibited symptoms of autism-related disorder.

She says that at the time, she wondered why so many children here had such symptoms, but as she did research she realized that every state in the country has had significant increases of cases of the disorder that were diagnosed.

To be classified as having autism spectrum disorder, a child has to show problems in three cognitive or developmental areas before the age of 3, Fry Williams says. The first set of criteria involve a child's ability to interact socially with others, which can be evidenced by lack of eye contact or inability to be comforted by being held.

A second factor is a delay or disorder in language development. Fry Williams says that about one-third of children with autism spectrum disorder begin developing language normally and then lose those skills, while others who have mild levels of disorder might have a very literal understanding of language or be unable to discern meaning from body language, Fry Williams says.

The third area is referred to as "restricted interest," which in severe cases might be expressed as repeating an action over and over, while in milder cases it could mean a singular preoccupation with any given subject matter.

People who have ASD often develop other concurrent disabilities, such as seizure disorders, attention deficit disorder, difficulty eating or hypersensitivity to external stimuli, and schizophrenia.

If problems in any of those three cognitive or developmental areas are noted, a child then would either be evaluated again by a physician after a period of time or referred to a specialist, generally a pediatric psychologist or pediatric psychiatrist, for further testing, Fry Williams says.

Once a child is diagnosed, parents are faced with a multitude of possible treatment and therapy options, ranging from swimming with dolphins, which some people believe has helped those with autism spectrum disorder, to vitamin supplements or dietary regimens.

Applying medical informatics

The two professors have envisioned a software system that uses what's called medical informatics, in which computer science is used to store, retrieve, and analyze medical data. Using artificial intelligence along with probability theory, the software will be able to analyze a large number of factors and apply different weights to each factor, they say.

The first software component will help with preliminary screening of children. With the software, they say, doctors will be able to use a sliding scale to enter the severity of various symptoms, and with that data, the software will evaluate whether the child is likely developing normally, should be rechecked in six months, or should be screened further.

The professors are preparing to submit a request for a National Institutes of Health grant in January to fund development of a second piece of the software that would help with more in-depth screening by utilizing tools developed by professionals and weighting different factors utilized by the tools. In that software component also lies the potential for research to define what symptoms are the most reliable in reaching a diagnosis, Fry Williams says.

They also have developed a prototype of a third piece of software, though more testing and refinement is needed on it, Mabry says. Its function will be to help monitor the treatments that a patient receives for autism spectrum disorder and possibly to help determine which therapies seem to be the most useful for different aspects of the disorder, they say. Called the Autism Spectrum Informatics Tool, it will allow physicians and possibly parents to collect information about a patient's treatments and how the patient responds to them, ultimately seeking to isolate which treatments or combination of treatments yield the best results.

As the professors continue to develop the system, they anticipate a need to access electronic health-record databases, such as one at the National Institutes of Health and the Kennedy Krieger Institute, based in Baltimore, Md., for which parents of children with autism spectrum disorder have provided information about treatments. More than 500 different treatments, for example, have been noted in the Kennedy Krieger database.

"These emerging databases make it possible for computer science to come in" and seek to analyze data provided by families living with autism spectrum disorder, Mabry says. "They don't know what to capture, so they are capturing everything."

Mabry and Fry Williams say they are looking into teaming up with clinical specialists to develop the system further.

"By integrating these tools, we're trying to respond to the demand with early identification, rapid intervention, and ongoing treatment evaluation," Mabry says.