Analyzing infant cries to detect autism shows potential, but definitive evidence is still lacking

A recent study published in the Journal of Autism and Developmental Disorders suggests that analyzing infant cries may hold promise for early detection of autism, but conclusive evidence is still lacking. Researchers conducted a systematic review and meta-analysis of existing studies examining the acoustic characteristics of autistic infants' cries.

While some studies reported consistent patterns in the cries of autistic infants, a pooled meta-analysis did not find statistically significant evidence for differences in pitch. The fundamental frequency (basic pitch) of autistic infants' cries showed a potential indication of being higher, but this finding was not supported when all studies were combined.

The paper highlights the importance of earlier identification of autism, as it can provide children and families with faster access to support during crucial periods of brain development. Current diagnosis methods rely heavily on clinical observation and assessment, which can be challenging to access in some communities. Researchers are exploring new approaches, including analyzing infant cries, as a potential early biomarker for autism.

Some studies have reported differences in cry features, such as pitch, duration, and variability, between autistic and non-autistic infants. Machine learning models have shown promising results, often achieving accuracies exceeding 90% in distinguishing between the cries of autistic and non-autistic infants. However, the study emphasizes the need for larger, more diverse, and longitudinal studies to refine machine learning models, establish standardized datasets, and better understand the developmental trajectories of acoustic markers.