Signal processing and classification for identification of clinically important parameters during neonatal resuscitation

Urdal, Jarle; Engan, Kjersti; Eftestøl, Trygve; Kidanto, Hussein; Yarrot, Ladislaus Blacy; Eilevstjønn, Joar; Ersdal, Hege

doi:10.1109/icsipa.2017.8120672

Cited by 2 publications

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…al. [19] signal processing and machine learning is utilized to find important parameters for survival after newborn resuscitation. Here it is shown that in addition to clinical parameters as birth weight and gestational age, the therapeutic strategies, i.e.…”

mentioning

confidence: 99%

Analysis and Knowledge Extraction of Newborn Resuscitation Activities from Annotation Files

Abukmeil,

Meinich-Bache,

Eftestøl

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Deprivation of oxygen in a newbornduring and after birth leads to birth asphyxia, which is considered one of the leading causes of death in the neonatal period. Adequate resuscitation activities are performed immediately after birth to save the majority of newborns. The primary resuscitation activities include ventilation, stimulation, drying, suction, and chest compression. While resuscitation guidelines exist, little research has been conducted on measured resuscitation episodes. Modeling the executed resuscitation activities to generate temporal data and extract knowledge can provide unique insights into dominant resuscitation activities. It also aids in constructing a resuscitation timeline to visually represent and describe the actions performed on a newborn.In this paper, we propose a method for generating and encoding temporal resuscitation data, enabling the description and visualization of the resuscitation timeline. We utilize neighborhood component analysis (NCA) to cluster the generated data based on the presence of ventilation and the outcome of the newborn. Additionally, we employ an autoencoder (AE) model to enhance clustering performance by visualizing its latent space.Our proposed method demonstrates high-quality visual clustering results on two different datasets. It provides insights into the intricate structure of the generated resuscitation data by grouping similar unlabeled resuscitation episodes into coherent clusters.

show abstract

mentioning

confidence: 99%