NewbornTime

NewbornTime  

Saving the Lives of At-Risk Newborns

NewbornTime

Saving the Lives of At-Risk Newborns

NewbornTime project at the University of Stavanger uses MOBOTIX thermal technology and artificial intelligence

The University of Stavanger (UiS) is an international university with 12,500 students and 2,200 employees. The UiS is organized into six academic faculties and several nationalresearch and competence centers. Education and research are carried out at the Facultyof Health Sciences. One research project is the NewbornTime project, which deals withinfant mortality. If an infant receives too little oxygen during and after birth, this can lead to asphyxia and the death of the baby. A newborn who needs help breathing must be resuscitated as quickly as possible. In Norway, approximately 10% of term infants need stimulation and around 3% need bag-mask ventilation.

CHALLENGE

The project partners – University of Stavanger (UiS), Stavanger University Hospital (SUS), Laerdal Medical, and bitUnitor – are working to automatically and instantly recognize a newborn as soon as it is born. What started with a portable thermal device has evolved into the NewBornTime concept. In the research project, a timeline is created that describes events and activities during a birth. The exact body temperature of the baby is not important. It is about the exact time of birth, which is to be determined automatically and in privacy, without attaching sensors to the mother or the baby. Manual determination of the periods is too imprecise. A video camera is a suitable sensor for this task.

SOLUTION

The exact time of birth to the second is determined automatically with the help of MOBOTIX thermal videos from the delivery room. Deep convolutional neural networks (CNNs) recognize the newborn, movements and resuscitation activities on thermal videos and convert them into AI models. Based on the resulting timelines, guidelines can be followed and evaluated, thus enabling successful resuscitation measures.

WHY MOBOTIX?

The project team chose MOBOTIX because the parties involved were looking for top quality from Europe and had already been using MOBOTIX systems reliably for many years. The accuracy of the temperature readings from the sensors was also impressive. Last but not least, the MOBOTIX IoT cameras could be integrated into the video management software (VMS), easily customized, send data to different file storage locations, and send alarms. There were no comparable systems in the same class that could have fulfilled this task so precisely.

 

CHALLENGE

In the research project, a timeline is created that describes events and activities during a birth. The exact body temperature of the baby is not important. It is about the exact time of birth, which is to be determined automatically and in privacy, without attaching sensors to the mother or the baby. Manual determination of the periods is too imprecise. A video camera is a suitable sensor for this task.

SOLUTION

The exact time of birth to the second is determined automatically with the help of MOBOTIX thermal videos from the delivery room. Deep convolutional neural networks (CNNs) recognize the newborn, movements and resuscitation activities on thermal videos and convert them into AI models. Based on the resulting timelines, guidelines can be followed and evaluated, thus enabling successful resuscitation measures.

WHY MOBOTIX?

The project team chose MOBOTIX because the parties involved were looking for top quality from Europe and had already been using MOBOTIX systems reliably for many years. The accuracy of the temperature readings from the sensors was also impressive.

 

CONCLUSION

The research findings offer a clear roadmap for specific actions and plans that can make a tangible difference in the lives of newborns. Implementing these insights will foster a more secure and healthier environment for newborns, ultimately leading to a brighter future for the next generation.

 

GDPR-compliant

Thermal cameras offer two advantages: they are GDPR-compliant and protect privacy as they do not show real images. They also reliably recognize the newborn, as the baby is warmer than the skin of the other people in the room at birth. One MOBOTIX Thermal camera is used in each delivery room, mounted on the ceiling above the head of the bed. The research team uses the thermal images to generate data to train AI models. It is important to know when resuscitation measures take place concerning the time of birth and what happens during resuscitation.

We informed the staff and parents how the system works and that the MOBOTIX thermal systems are GDPR-compliant. With their signature, the parents gave their permission to use the data from the MOBOTIX thermal sensors for the study. When a significant majority – over 80% – of mothers surveyed expressed excitement about joining the study, we knew right away – this is something special. We’re proud to be on this journey together!

James Nathan Miller, Bravida Managing Director

Learn more about the NewbornTime project

Download the full case study to gain detailed insights into this innovative healthcare solution and learn how AI-powered camera monitoring can help improve neonatal resuscitation and prevent birth asphyxia in your hospital.

Case Study NewbornTime (PDF, 1.82 MB)
| EN | DE | ES | FR | IT