This is a preliminary study whose objectives are to define the clinical use cases and the constraints of the implementation of a multi-sensor image-sound system.

300 000 newborns per year are hospitalized in European neonatal units in the post-natal period. This is a critical period for neurobehavioral development and the establishment of an optimal model of sleep-wake cycle is fundamental for brain development.

The objective of the research is to develop new tools for computer-aided clinical and predictive diagnosis and real-time monitoring in neonatology.

The development of non-invasive sensors based on video and sound images aims to improve the short- and medium-term prognosis in perinatal patients.

To that aim, multi-modal composite indices integrating the image (which is used only to generate a quantification of the movement) and the sound to the bioelectric signals already available (electrocardiogram, oxymetry, temperature) will be developed to provide diagnostic and prognostic information.

Inclusion Criteria:

• newborns less than 6 weeks old in corrected terms;

• newborns in one of the following situations:

  • persistent discomfort (apneas);
  • infection or suspicion of infection;
  • premature with uncomplicated evolution;
  • Premature brain lesion acquired in perinatal;
  • newborn having been exposed to premature rupture of the membranes.

Exclusion Criteria:

• Congenital malformation;
• Metabolic disease.

This study is aimed at newborns hospitalized at the CHU in Rennes