1st non invasive real-time detection by AI of infectious risk in premature newborns

First sepsis (infectious) risk scores were successfully obtained during a non-invasive 12h-recording of premature newborns at Rennes University Hospital.

Digi-NewB system (screenshot) © Digi-NewB
  1. DigiNewb system : an innovative tool for an earlier detection of septis on premature newborns
  2. How does Digi-NewB work?
  3. A great step made possible by the common work of a National and European Network

This new generation monitoring system aims to save lives and improve health and care for the 300 000 babies born prematurely each year in Europe. Three years after the start of the Digi-NewB project, the Digi-NewB European consortium is now ready to show a first demonstration of its AI-based sepsis risk decision-support system.

DigiNewb system : an innovative tool for an earlier detection of septis on premature newborns

Prof.Pladys - © CHU Rennes

    We are often ‘too late’ when we detect sepsis, at the peak of clinical signs. Each hour we can gain is an expected benefit for the patient. With our approach we both want to demonstrate that we can early detect sepsis several hours to days before and be more specific than usual practice,

says Pr. Pladys, Digi-NewB Scientific coordinator, Head of the Pediatric Unit of Rennes University Hospital, Faculty of Medicine Professor at the Université de Rennes 1.

How does Digi-NewB work?

The Digi-NewB system consists of three key steps:

  •  capturing a full set of data from the newborn in a non-invasive way (vital signs monitoring, clinical observations, movements thanks to video recordings by IR and HD cameras) ;
  • treating and analyzing the data via artificial intelligence algorithms ;
  • displaying a newborn sepsis risk score.

 Thanks to this new system, clinicians can view a sepsis risk score evolution in time, view selected parameters trends, and replay video recordings for any given time. The system is easy to use and install, thanks to usability studies held in several hospitals with care staffs.

Digi-NewB is based on data collected with the parents' consent of 400 newborns in the care of the Pediatric Units of 6 University Hospitals in Western France (HUGO network - Rennes, Angers, Brest, Nantes, Tours, Poitiers).
It makes extensive use of the Laboratory for Signal and Image Processing (LTSI, Université de Rennes 1/Inserm) expertise in cardiac and respiratory variabilities, Big Data, AI, computer modelling, and database administration.

To validate the system, Digi-NewB leaders aim to launch a large clinical involving more than 1,000 newborns in additional hospitals.
 

A great step made possible by the common work of a National and European Network

Digi-NewB was developed by a European consortium made of 7 academic and industrial partners funded by the EU H2020 research and innovation program.

These partners collaborate with the GCS HUGO network, Rennes Hospital and the Université de Rennes 1 on:

  • user-centered design (NUIG, Irland)
  • Artificial intelligence models (TUNI, Finland)
  • complexity indexes (INESC TEC, Portugal)
  • hardware development, sound and video analysis (Voxygen, France)
  • innovative hospital information systems (Syncrophi, Ireland)