The article is available in another language

Can discover damage to premature born children’s brains after just a few months

The article is available in another language

Clinicians and scientists at St. Olavs hospital and the Norwegian University of Science and Technology (NTNU) are now able to detect brain damage and Cerebral Palsy through observing newborn babies' spontaneous movements. In the project “In-Motion”, video recordings are sent to experts with specialist experience within clinical evaluations. In addition, machine learning algorithms and models for automatic tracking of Cerebral Palsy are under development. CheckWare plays a key role in transmitting the videos.

Around 120 babies and toddlers are affected by Cerebral Palsy (CP) annually in Norway. These are often babies who are prematurely born, experience lack of oxygen during birth, or who have had a brain hemorrhage.

Some children with CP are detected during follow-up at the hospital, while others are detected through primary care. As it is easier to observe the disability when the child starts crawling, standing up and walking, most children get their diagnosis when they are 1-2 years of age.

InMotion

Cooperation between St. Olavs hospital, Central Norway RHA and the Norwegian University of Science and Technology

Current capacity of 50-60 children in Trøndelag county.

The sooner the better

Adde has more than 22 years of experience working with diagnostics and treatment at the Neonatal Intensive Care Unit, as well as follow-up programs of sick newborns in the specialist health service. He has a background as a physiotherapist specialized in pediatric physiotherapy, as well as a PhD in clinical medicine.

He cannot express the importance of examining and assessing the newborns early. In advanced CP, there is a risk of other complications such as joint stiffness, pain, nutritional difficulties, impaired vision and hearing, which in turn can lead to repeated hospitalizations and surgeries.

It is important to discover who has the risk of getting the condition early, but it is equally important that we find out who does not have, nor the risk for developing, CP. The fact that 8-10 percent of children in the medical risk group get CP also means that around 90 per cent do not get CP.

Babies are recorded at home in normal, familiar conditions

The clinical observation method analyzing the baby’s movements is called General Movement Assessment (GMA) and is used when the baby has passed its original due date by 2-4 months. The method is recommended to detect the children with a high risk of developing CP. By using this method, it can be determined whether the child is at high or low risk of developing CP.

GMA was developed early in the 2000’s and is comprehensively documented to this day. The child is video recorded while lying on its back while moving undisturbed, in order to observe the child’s spontaneous movements. The recording lasts for 3 minutes. Afterwards, the video can be analyzed using GMA and watched by trained clinical observers.

- For many years, the evaluation was only available at the hospital. In order to examine the child, both the child and the parents had to meet at the hospital for an appointment. Conducting the observation required several variables to be in place: the child had to be awake, active, comfortable, and feeling well when doing the video recording. Today, the video is recorded at home when it suits the child best. The parents now get to record the video in peace and quiet without having to travel to the hospital, Lars Adde explains.

He adds:

- Not only does this contribute to a higher quality on the video, but it also contributes to including the parents more in the process, instead of being passive bystanders.

The parents upload the video in CheckWare, and the clinician at the hospital is able to access the video right away. The GMA-evaluation is conducted remotely, and the family is notified about the process, including check-ups at the hospital. Thus, it is possible to detect symptoms of CP at an earlier stage.

For many years, the evaluation was only available at the hospital. In order to examine the child, both the child and the parents had to meet at the hospital for an appointment. Conducting the observation required several variables to be in place: the child had to be awake, active, comfortable, and feeling well when doing the video recording. Today, the video is recorded at home when it suits the child best. The parents now get to record the video in peace and quiet without having to travel to the hospital. Not only does this contribute to a higher quality on the video, but it also contributes to including the parents more in the process, instead of being passive bystanders.

Lars Adde

Specialist in pediatric physiotherapy and responsible for all medical research activities, In-Motion

CheckWare with video solution for in-home use

Choosing the right solution for remote monitoring has been immensely important for the In-Motion project, especially because of the privacy reasons. There were challenges related to using smart phones for recording videos of the patients, and the project therefore had to scour the market for a solution that complied with all the privacy regulations.

- We have been working on various solutions for a long time, including a very costly app. Eventually, we came across CheckWare, which already had a system for sending out measures to patients. The solution was clarified and approved by the Central Norway RHA, which was an important factor to us, Adde explains.

Cerebral palsy

Cerebral palsy (CP) is the most common cause of motor impairment in children, and occurs in approximately two out of 1000 live births. (FHI.no)

120 Norwegian childrens gets the CP diagnosis annually (FHI.no).

There is no cure against CP, but treatment can reduce pains and improve motor skills (FHI.no).

HEMIT (Central Norway RHA IT), which is the IT-department at St. Olavs hospital, was asked to work on a custom-built solution for submitting video files from parents with premature born children. In collaboration with CheckWare, a system was created: It contains automatic messaging when using text messages to parents, and possibilities of providing parents with learning material among other things.

- The parents involved use the CheckWare solution to upload the videos for us to analyze. We get a unique overview of who has received messages, submitted videos, received reminders and so on. For those of us collecting videos, this is revolutionary. Previously, we have used Excel-sheets and manual processes including phones and Microsoft Outlook. Now, we have a fully automated solution in CheckWare, Adde says.

He emphasizes that they are able to conduct thorough quality controls. Through measures and questionnaires, they make sure that parents experience the method to be a safe way to record and transmit videos of their children.

- The parents consider this to be a great way of submitting the video. It is perceived as secure and feasible, and just as secure as conducting the recording at the hospital.

He continues:

- I talk about and compliment the CheckWare solution all the time, even though we started this only 6 months ago, he says, and would like to emphasize on a key person at HEMIT:

- Per Henning Valderhaug is an advisor at HEMIT. He has done a great job at implementing the solution, and he has been the link between the clinical environment and CheckWare. He should definitely be able to bask in the glory, as he has been crucial to this project, Adde believes.

The parents involved use the CheckWare solution to upload the videos for us to analyze. We get a unique overview of who has received messages, submitted videos, received reminders and so on. For those of us collecting videos, this is revolutionary. Previously, we have used Excel-sheets and manual processes including phones and Microsoft Outlook. Now, we have a fully automated solution in CheckWare

Lars Adde

Specialist in pediatric physiotherapy and responsible for all medical research activities, In-Motion

Machine vs human

Behind In-Motion is an interdisciplinary research team consisting of physiotherapists, neonatal doctors, pediatricians, motion scientists and computer engineers. The In-Motion team covers national and international clinical studies at multiple locations in Norway, Denmark, the United States, Belgium, China, India, and Turkey. Together they form a database for multicenter studies.

Adde is a specialist in Pediatric Physiotherapy, a GMA observer and knows everything about the spontaneous movements of infants. With him on the team is also Espen A. F. Ihlen, who is responsible for the machine learning algorithm together with Daniel Groos at NTNU. Groos defended his dissertation only a few weeks ago. Having an academic environment such as Norwegian Open AI-lab, with professor Heri Ramampiaro at the forefront, has been crucial. This environment will play a central role in the further development of automation and digitalization in In-Motion.

Worthy of note is that In-Motion is based on two “pillars”: Pillar one uses medical expertise to classify spontaneous movements, and risk assesses the probability of developing CP. The second pillar uses machine learning models to do the same, but with automatic image analysis. Both pillars use 3-minute video footage of the infant's spontaneous movements at 2-4 months of age as a basis.

-We train a machine learning algorithm to recognize movement patterns in a video recording. Then we use this to try to predict the likelihood of the child developing CP. The automated results will only be used as decision support in the decision-making processes that are still undertaken by human expertise, says Adde.

He explains that the processes are quite complicated and that the brain of the expert must be trained in order to recognize the movement patterns of a baby. The competence should be maintained, while they at the same time are working on recruiting new experts.

The dream of an automated motion analysis

In-Motion currently follows up on 50-60 children annually within the Central Norway Regional Health Authority. On a national basis, they envision being able to follow up on 7-800 children a year.

- I have had this dream of video recording and automatics movement analysis for 20 years. Remotely based early risk assessment of all sick newborns in Norway is actually within reach. This is because we have found a solution where we have been able to retrieve the videos in a good and patient-safe way, he says, and explains that the video solution is not yet available in other regions than the Central Norway RHA for now.

Lars Adde is a part of a service innovation project in Central Norway RHA together with Beate Horsberg Eriksen at Ålesund Hospital, Kristin Åberg at Levanger Hospital and Kristine Grünewaldt at St. Olavs hospital.

- We are now working towards being able to use the solution several places in Norway. We want to extend the CheckWare solution to other RHA’s for remotely based expert evaluations.

- The ultimate dream is for the whole world to have access, but first we need to test the algorithm in a pilot. Initially, it is realistic to work towards being able to offer this service in all of the hospitals in Norway. We need to make sure that as many new families as possible have access to in-home evaluations as soon as possible.

In-Motions use of CheckWare

Respondent tools:

Measures
Plans

Clinician tools:

Clinical reports
Clinician insight
Treatment roles
Raw data exports

System management tools:

System settings
Access control
System log
System reports
Respondent management
Clinician management
Authentication: Bank ID (Norway) / security clearance level 4