Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks
The monitoring of vital signs and increasing patient comfort are cornerstones of modern neonatal intensive care. Commonly used monitoring methods are based on skin contact which can cause irritations and discomfort in preterm neonates. Therefore, non-contact approaches are the subject of current res...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/10/4910 |
| _version_ | 1827740121706790912 |
|---|---|
| author | Johanna Gleichauf Lukas Hennemann Fabian B. Fahlbusch Oliver Hofmann Christine Niebler Alexander Koelpin |
| author_facet | Johanna Gleichauf Lukas Hennemann Fabian B. Fahlbusch Oliver Hofmann Christine Niebler Alexander Koelpin |
| author_sort | Johanna Gleichauf |
| collection | DOAJ |
| description | The monitoring of vital signs and increasing patient comfort are cornerstones of modern neonatal intensive care. Commonly used monitoring methods are based on skin contact which can cause irritations and discomfort in preterm neonates. Therefore, non-contact approaches are the subject of current research aiming to resolve this dichotomy. Robust neonatal face detection is essential for the reliable detection of heart rate, respiratory rate and body temperature. While solutions for adult face detection are established, the unique neonatal proportions require a tailored approach. Additionally, sufficient open-source data of neonates on the NICU is lacking. We set out to train neural networks with the thermal-RGB-fusion data of neonates. We propose a novel indirect fusion approach including the sensor fusion of a thermal and RGB camera based on a 3D time-of-flight (ToF) camera. Unlike other approaches, this method is tailored for close distances encountered in neonatal incubators. Two neural networks were used with the fusion data and compared to RGB and thermal networks. For the class “head” we reached average precision values of 0.9958 (RetinaNet) and 0.9455 (YOLOv3) for the fusion data. Compared with the literature, similar precision was achieved, but we are the first to train a neural network with fusion data of neonates. The advantage of this approach is in calculating the detection area directly from the fusion image for the RGB and thermal modality. This increases data efficiency by 66%. Our results will facilitate the future development of non-contact monitoring to further improve the standard of care for preterm neonates. |
| first_indexed | 2024-03-11T03:20:31Z |
| format | Article |
| id | doaj.art-898e59b16aa54e67aaa82035e4fb19a4 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T03:20:31Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-898e59b16aa54e67aaa82035e4fb19a42023-11-18T03:14:30ZengMDPI AGSensors1424-82202023-05-012310491010.3390/s23104910Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural NetworksJohanna Gleichauf0Lukas Hennemann1Fabian B. Fahlbusch2Oliver Hofmann3Christine Niebler4Alexander Koelpin5Nuremberg Institute of Technology, 90489 Nuremberg, GermanyNuremberg Institute of Technology, 90489 Nuremberg, GermanyDivision of Neonatology and Pediatric Intensive Care, Department of Pediatrics and Adolescent Medicine, Friedrich-Alexander-Universität Erlangen-Nürnberg, 91054 Erlangen, GermanyNuremberg Institute of Technology, 90489 Nuremberg, GermanyNuremberg Institute of Technology, 90489 Nuremberg, GermanyHamburg University of Technology, 21073 Hamburg, GermanyThe monitoring of vital signs and increasing patient comfort are cornerstones of modern neonatal intensive care. Commonly used monitoring methods are based on skin contact which can cause irritations and discomfort in preterm neonates. Therefore, non-contact approaches are the subject of current research aiming to resolve this dichotomy. Robust neonatal face detection is essential for the reliable detection of heart rate, respiratory rate and body temperature. While solutions for adult face detection are established, the unique neonatal proportions require a tailored approach. Additionally, sufficient open-source data of neonates on the NICU is lacking. We set out to train neural networks with the thermal-RGB-fusion data of neonates. We propose a novel indirect fusion approach including the sensor fusion of a thermal and RGB camera based on a 3D time-of-flight (ToF) camera. Unlike other approaches, this method is tailored for close distances encountered in neonatal incubators. Two neural networks were used with the fusion data and compared to RGB and thermal networks. For the class “head” we reached average precision values of 0.9958 (RetinaNet) and 0.9455 (YOLOv3) for the fusion data. Compared with the literature, similar precision was achieved, but we are the first to train a neural network with fusion data of neonates. The advantage of this approach is in calculating the detection area directly from the fusion image for the RGB and thermal modality. This increases data efficiency by 66%. Our results will facilitate the future development of non-contact monitoring to further improve the standard of care for preterm neonates.https://www.mdpi.com/1424-8220/23/10/4910non-contact monitoringneonatessensor fusionneural networkface detection |
| spellingShingle | Johanna Gleichauf Lukas Hennemann Fabian B. Fahlbusch Oliver Hofmann Christine Niebler Alexander Koelpin Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks Sensors non-contact monitoring neonates sensor fusion neural network face detection |
| title | Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks |
| title_full | Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks |
| title_fullStr | Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks |
| title_full_unstemmed | Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks |
| title_short | Sensor Fusion for the Robust Detection of Facial Regions of Neonates Using Neural Networks |
| title_sort | sensor fusion for the robust detection of facial regions of neonates using neural networks |
| topic | non-contact monitoring neonates sensor fusion neural network face detection |
| url | https://www.mdpi.com/1424-8220/23/10/4910 |
| work_keys_str_mv | AT johannagleichauf sensorfusionfortherobustdetectionoffacialregionsofneonatesusingneuralnetworks AT lukashennemann sensorfusionfortherobustdetectionoffacialregionsofneonatesusingneuralnetworks AT fabianbfahlbusch sensorfusionfortherobustdetectionoffacialregionsofneonatesusingneuralnetworks AT oliverhofmann sensorfusionfortherobustdetectionoffacialregionsofneonatesusingneuralnetworks AT christineniebler sensorfusionfortherobustdetectionoffacialregionsofneonatesusingneuralnetworks AT alexanderkoelpin sensorfusionfortherobustdetectionoffacialregionsofneonatesusingneuralnetworks |