Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

FEM-based volume estimation using electrode catheter measurements.

Journal of electrical bioimpedance·2026
Same author

For a better understanding of thermoregulation in premature infants≤28 weeks of gestation: infrared thermography (IRT) and photoplethysmography imaging (PPGI) in a hybrid camera model as non-contact measuring methods.

Zeitschrift fur Geburtshilfe und Neonatologie·2026
Same author

A systematic review of pandemic ventilator designs.

Biomedizinische Technik. Biomedical engineering·2026
Same author

Spatial temperature monitoring of preterm infants using a multi-modal camera setup.

Medical & biological engineering & computing·2026
Same author

Non-contact monitoring of vital parameters in preterm infants: Insights from infrared thermography and photoplethysmography imaging.

Zeitschrift fur Geburtshilfe und Neonatologie·2026
Same author

Non-contact monitoring in neonatology.

Zeitschrift fur Geburtshilfe und Neonatologie·2026

Related Experiment Video

Updated: Aug 6, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K

Multi-modal body part segmentation of infants using deep learning.

Florian Voss1, Noah Brechmann2,3, Simon Lyra2

  • 1Chair of Medical Information Technology, Helmholtz Institute for Biomedical Engineering, RWTH Aachen University, Aachen, Deutschland. voss@hia.rwth-aachen.de.

Biomedical Engineering Online
|March 23, 2023
PubMed
Summary
This summary is machine-generated.

Deep learning models accurately segment infant body parts from thermal and visible light images. Feature fusion achieved the best results, improving temperature monitoring for premature infants.

Keywords:
Body part segmentationDeep learningInfrared thermographyNICUNeonatal intensive careSemantic segmentation

More Related Videos

A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants
11:14

A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants

Published on: October 4, 2015

11.0K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

470

Related Experiment Videos

Last Updated: Aug 6, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K
A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants
11:14

A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants

Published on: October 4, 2015

11.0K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

470

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Neonatal Care

Background:

  • Monitoring premature infant body temperature is crucial for optimal care and early disease detection.
  • Thermography offers a non-contact, wireless alternative to traditional cable-based temperature monitoring methods.
  • Automatic segmentation of infant body regions is essential for clinical thermography due to infant movement.

Purpose of the Study:

  • To develop and evaluate deep learning algorithms for automatic infant body part segmentation.
  • To compare the performance of U-Net based neural networks using visible light, thermography, and fused modalities.
  • To assess the impact of transfer learning and data augmentation on segmentation accuracy.

Main Methods:

  • Developed three U-Net based deep learning models: one for visible light, one for thermography, and one for feature fusion.
  • Created and manually labeled a dataset of 600 visible light and 600 thermography images from 20 infant recordings.
  • Utilized transfer learning on adult datasets and data augmentation to enhance segmentation performance.

Main Results:

  • Transfer learning and data augmentation improved segmentation accuracy across all imaging modalities.
  • The feature fusion model achieved the highest mean Intersection-over-Union (mIoU) of 0.85, followed closely by the visible light (RGB) model.
  • The thermography-only model achieved a lower mIoU of 0.75; segmentation accuracy was reduced for the torso when only small skin areas were visible.

Conclusions:

  • Multi-modal neural networks offer a novel approach for infant body segmentation, particularly with limited data.
  • Feature fusion, cross-modality transfer learning, and data augmentation yield robust segmentation results.
  • This approach enhances the potential for non-contact, wireless temperature monitoring in neonates.