Early health diagnoses from a single baby cry

Brown University scientists have developed a baby crying analysis system that can help diagnose illnesses from the acoustic quality of the cry alone.

When a baby cries, it’s usually because they need something — whether that’s breast-feeding, changing, burping or simply attention. But according to researchers at Brown University, babies cries also contain extra information that indicates the state of their health. The scientists have now developed a baby crying analysis system that can help diagnose illnesses from the acoustic quality of the cry alone.

A collaboration between the university and the Women & Infants Hospital in Rhode Island, the research showed that a range of different conditions manifest themselves acoustically in the cry of babies, although the difference is imperceptible to human ears. Its digital platform takes recordings of cries, splits them into small 12.5-millisecond clips before analyzing them for any of 80 flags that may indicate anything from neurological problems or developmental disorders. The flags in each clip are then averaged out over the duration of a single cry to determine which ones are most pertinent. Harvey Silverman, professor of engineering and director of Brown’s Laboratory for Engineering Man/Machine Systems: “It’s a comprehensive tool for getting as much important stuff out of a baby cry that we can.” The idea is that the system could detect early signs of disorders such as autism, which typically aren’t apparent until the child is older.

Although the researchers don’t currently have a plan in mind to bring the system to market, it’s easy to see how an app could be developed. Could other kinds of health information be gleaned from unusual sources such as this?

Spotted by Murtaza Patel, written by Springwise