The system evaluates 80 different parameters altogether, each having the potential to hold clues about a baby's health.
To a parent, a baby’s’ cry could mean they’re hungry, sleepy, in pain or uncomfortable. To scientists, the subtle acoustic characteristics of those cries, often unnoticeable to the human ear, might contain important information about a child’s health.
Researchers from Brown University and Women & Infants Hospital of Rhode Island have developed a new computer based tool that can perform finely tuned acoustic analysis of a baby’s cry. The research team hopes the baby cry analyzer will lead to new ways both researchers and clinicians can identify neurological problems and development disorders in children through their cries.
Stephen Sheinkopf, assistant professor of psychiatry and human behavior at Brown University, helped design the computer based tool. In an interview he said, “There are lots of conditions that might manifest in differences in cry acoustics. For instance, babies with birth trauma or brain injury as a result of complications in pregnancy or birth or babies who are extremely premature can have ongoing medical effects. Cry analysis can be a noninvasive way to get a measurement of these disruptions in the neurobiological and neurobehavioral systems in very young babies.”
This new analyzing tool is the result of two years of collaborative work between the faculty at Brown’s School of Engineering and the hospital based faculty at Women & Infants Hospital.
The analysis takes place in two different phases. In the first phase, the analyzer separates a recorded cry into 12.5 millisecond frames. Each frame is then analyzed for a variety of different parameters. Those parameters include frequency characteristics, voicing and acoustic volume.
During the second phase, the data gathered during the first phase is further analyzed to provide a broader view of the cry, reducing the number of parameters down to the most useful ones. The frames are put back together and characterized as an utterance, which is a single “wah” sound, or silence which is the pause between the utterances. The longer utterances are separated from the shorter ones, and the time between each is recorded. The pitch, as well as the contour of the pitch and many other variables, can then be averaged across each utterance. The system evaluates 80 different parameters altogether, each having the potential to hold clues about a baby’s health.
Harvey Silverman, professor of engineering and director of Brown’s Laboratory for Engineering Man/Machine Systems, says,”It’s a comprehensive tool for getting as much important stuff out of a baby cry that we can.”
Silverman and his graduate students Brian Reggiannini and Xiaoxue Li worked with Sheinkopf and Barry Lester, director of Brown’s Center for the Study of Children at Risk, to understand what important things they needed to be looking for in a baby’s cry.
Lester, who has studied baby cries for several years, says this new research relates back to the 60s and a disorder referred to as Cri du chat syndrome. Cri du chat is the result of a genetic anomaly similar to Downs Syndrome and is characterized by a distinct, high pitched cry that sounds similar to a cat.
“The idea is that cry can be a window into the brain,” says Lester. “Cry is an early warning sign that can be used in the context of looking at the whole baby.”
A paper describing the new tool is featured in the Journal of Speech, Language and Hearing Research.