In April, 2021, the 4TU.NIRICT research centre launched the ‘Research Impulse 2021’ fund to support already running ICT research activities or new (planned) activities that could benefit from start-up funding. Applications could be submitted until May 1, 2021.
4TU.NIRICT received 14 proposals in total with a good spread of the 4TUs. After thoroughly reviewing all proposals by the members of the NIRICT Board, two projects have been awarded funding. The first project focuses on measures of breastfeeding and infant sucking behavior and the second one on robots for pain management in children.
FLOW: Measures of breastfeeding and infant sucking behavior
- Dr. ir. J.A.M. Haarman, Postdoctoral researcher - Human Media Interaction, Faculty of
Electrical Engineering Mathematics and Computer Science, University of Twente
- Dr. E.M. Brouwer-Brolsma, Assistant professor - Division of Human Nutrition and
Health- global nutrition, Wageningen University and Research.
- Dr. M.P. Lasschuijt, Postdoctoral researcher - Division of Human Nutrition and Health -
Sensory Science and Eating Behavior, Wageningen University and Research.
- Dr. I. Kalinauskaite, Postdoctoral researcher - Department of industrial Design,
Systemic Change group, Eindhoven University of Technology.
Sufficient milk intake by newborns is important. Not all infants are able to self-regulate the amount of breastfeeding that is needed, while sufficient nutrient intake is crucial, especially in the first 6 months. Current breastfeeding assessment methods are invasive, or inaccurate and may affect normal feeding practices.
The aim of this proposal is to take the first steps in developing a breastfeeding assessment device that is able to measure direct milk intake and sucking behavior of infants in a non-invasive manner. It aims to transform sensing principles into wearable, non-intrusive assessment devices (by integration in attributes such as bibs, hairbands, or skin stickers). Different assessment methods will be explored, such as motion sensors to detect swallow-movement, strain sensors to detect breast volume changes/swallow movements, or imaging techniques (such as laser-doppler) to assess milk-flow. A literature study combined with a stakeholder analysis will be done on how to best capture breastfeeding and sucking behavior. Based on this we will define a clear set of requirements. The most promising method will be further developed into a first wearable (rapid) prototype. Using the prototype, suitable analysis techniques will be explored to account for proper event detection using time series threshold analysis vs machine learning. Additionally, IoT integration is critical for ensuring non-obtrusive data collection. The development of such breastfeeding sensor will aid in monitoring milk intake of infants thereby contributing to infant health.
Robots for pain management in children
- Dr. Ir. Emilia I. Barakova, ID, TU Eindhoven (TU/e) (Main applicant)
- Dr. Ir. Edwin Dertien, assistant professor, FEEMCS, University of Twente (UT)
- Dr. Ir. Dennis Reidsma, associate professor, HMI, University of Twente
- Dr. Frank Broz, assistant professor, EEMCS, TU Delft (TUD)
- Additional members: Prof. Ignacio Malagon, Radboud UMC, Prof. Paula Sterkenburg,
This project will develop technological solutions that include wearables for measuring physiological signals related to sensation and fear of pain and child-robot interactive behaviors for pain detection and accurate diagnosis. AI algorithms will be used for both pain detection and selecting adequate robot behaviors.
Reducing children’s suffering from pain or fear of pain with non-pharmacological interventions can have many benefits. Pain distraction, self-exploration of pain in movement therapies, and explanation of predicted severity of pain are established as promising methods for managing pain associated with children’s medical procedures.
The project builds on two ongoing collaborations between TU/e-VU on pain and stress detection through physiological measurements and collaboration between TU Eindhoven, Twente University, and Radboud Medical Center on developing child-robot interactions for pain management. The over 10 years TU/e-VU collaboration resulted in developing the “Smart sock”, i.e., a sock with embedded sensors that connects to a mobile application to visualize the pain and stress-related signals, which facilitated two Ph. D. projects.
The 4TU funding will help develop robust and personalized prototypes, integrated system solutions, and novel child-robot interactions enhanced by combining physiological signals and human expertise in pain detection and management. The project will serve as a steppingstone for further collaboration and future projects between the mentioned universities.