Analyzing children’s voices using machine learning could aid early diagnosis of Rett syndromeexplains a small study.
Researchers say such audio analysis among 6- to 11-month-old babies may help identify signs of Rhett, or Fragile X syndromelong before a child is usually diagnosed with these childhood disorders.
“We… prove it, even if the individuals they were studied with were [fragile X] And the [Rett] They have not yet shown clinical signs in the second half of their life, the machine “hears” that they have already pronounced differently than TD [typically developing] individuals,” the researchers wrote.
The team noted that these early results were based on data from a small number of individuals, and stressed the need for further research to confirm and extend this approach.
the study, “Speech-based automatic detection of fragile X syndrome and Rett syndrome“in the magazine Scientific Reports.
Both Rett and Fragile X are developmental disorders that symptoms It usually begins in early childhood. Speech abnormalities are usually associated with both genetic conditions.
In Rett syndrome, as well as in fragile X syndrome, patients often have a delay in getting a correction diagnose. Many symptoms are common to a range of disorders, and are not sufficient alone to diagnose disease conditions.
Speech investigation for the diagnosis of Rett syndrome
Now, a team of scientists in Germany and Austria has tested the idea that automated analyzes of audio recordings may be useful in the early identification of Rett and the fragile X.
The study included three children with Rett syndrome, and three children with one fragile X. The children’s ages ranged from 6 to 11 months.
All children with Rett syndrome were female, in line with the prevalence of gender-based disorder, whereas all children with fragile X were male. A group of six typically developing children, matched for age and sex, were included as controls.
Quite simply, the analysis involved taking audio data from the children’s home recordings, then entering the data into a computer along with a set of mathematical rules. The computer will then use these rules to “learn” how to sort the children into pre-defined groups.
In an initial set of tests, the researchers evaluated whether these analyzes could differentiate children with Rett or Fragile X from sex-matched controls, or if they could differentiate between abnormal development (Rett or Fragile X) and typical development. In these early analyses, the computer performed with 100% accuracy.
These findings “not only demonstrate the basic feasibility, but indicate the high potential of a future practical application approach in pediatric health care,” the scientists wrote.
Moreover, closer examination of the data showed that the acoustic features that were important for making these differences in the computer models were distinct for Rett compared to the brittle X.
“This suggests that the early verbal characteristics of individuals with [fragile X] and individuals [Rett] It manifests phonetically in different ways compared to the early verbal behavior typical of gender-matched controls,” the researchers wrote.
In further analyses, investigators attempted to use phonemic analysis to sort all children into the appropriate group – Rett, fragile X, or control.
“The current study was the first attempt to combine early speech data of individuals with various recently discovered genetic disorders into a single classification model,” the team noted.
Of the twelve children, nine were correctly classified. One child with Fragile X was incorrectly classified as having Rett, another child with Fragile X and one child with Rett were incorrectly classified as developing typically.
The researchers identified some potential changes to the algorithm that could be useful in improving accuracy. They also stressed the importance of further research with larger data sets to validate and refine the approach.
“Although our findings suggest that this approach is [worthwhile] To be further followed, it must be interpreted carefully and cannot be generalized. The scientists wrote this primarily due to the extremely small data set.
A limitation was also noted in the fact that analyzes were based on recordings from home videos.
“It should be borne in mind that the recordings were not made originally with the parents’ intention to collect data for later scientific analyses, but usually to create a memory of family routines and special moments of their children’s childhood,” the researchers wrote, noting that this may lead to an underrepresentation of the abnormal behavior that chose Parents not to register it.
Despite these limitations, the researchers stated that the use of home video “provides a unique opportunity to study early development in a natural setting and currently represents the best available approach for objective investigation of initiators.” [early] Behavioral phenomena in rare, delayed developmental disorders such as [fragile X] or [Rett]. “