Elsevier

Early Human Development

Volume 89, Issue 12, December 2013, Pages 943-948
Early Human Development

Non-contact heart rate monitoring utilizing camera photoplethysmography in the neonatal intensive care unit — A pilot study

https://doi.org/10.1016/j.earlhumdev.2013.09.016Get rights and content

Abstract

Background

Presently the heart rate is monitored in the Neonatal Intensive Care Unit with contact sensors: electrocardiogram or pulse oximetry. These techniques can cause injuries and infections, particularly in very premature infants with fragile skin. Camera based plethysmography was recently demonstrated in adults as a contactless method to determine heart rate.

Aim

To investigate the feasibility of this technique for NICU patients and identify challenging conditions.

Study design and participants

Video recordings using only ambient light were made of 19 infants at two NICUs in California and The Netherlands. Heart rate can be derived from these recordings because each cardiovascular pulse wave induces minute pulsatile skin color changes, invisible to the eye but measurable with a camera.

Results

In all infants the heart beat induced photoplethysmographic signal was strong enough to be measured. Low ambient light level and infant motion prevented successful measurement from time to time.

Conclusions

Contactless heart rate monitoring by means of a camera using ambient light was demonstrated for the first time in the NICU population and appears feasible. Better hardware and improved algorithms are required to increase robustness.

Introduction

The heart rate (HR) is a critical parameter to assess the well-being of infants in the neonatal intensive care unit (NICU). Episodes of bradycardia occur frequently in premature infants due to a variety of causes, including immature respiratory development or sepsis [1], [2]. Current HR monitoring techniques include electrocardiography (ECG) and pulse oximetry based on photoplethysmography (PPG) [3]. These techniques are reliable and inexpensive. However, there are several disadvantages because both are contact devices using adhesive sensors. Placement and removal of the sensors and the attachment to wires cause discomfort, stress, pain and sometimes even epidermal stripping. The latter occurs particularly in infants < 27 weeks of gestational age, because in these infants the bond between the sensor and dermis is stronger than the epidermal–dermal junction [4], [5], [6]. The patches and wires appear on X-rays and complicate reading the film. Furthermore, the obtrusiveness of the wires impairs parent–child bonding, especially during kangaroo mother care. Anand and Scalzo suggested that pain, stress and maternal separation of NICU patients have a negative impact on cognitive development [7]. In addition, a study by Chen et al indicates that repetitive application and removal of patches, adversely affect the infant's well-being and developmental outcome [8]. In view of the above, a contactless HR monitoring technique would be highly desirable.

Recently, camera based PPG was demonstrated [9], [10], [11], [12], [13], [14], [15], [16], [17]. This novel technique is entirely contactless but uses the same principle as contact PPG, better known as pulse-oximetry. Blood absorbs light more than surrounding tissue so variations in blood volume affect light transmission and reflectance. While contact PPG predominantly uses transmissive mode, camera PPG is typically performed in reflective mode. The cardiovascular pulse waves cause changes in the volume of arterioles which result in minute pulsatile skin color changes. These color changes may be compared to blushing (more blood, redder skin), except that they are less intense. While invisible to the eye, these ‘micro-blushes’ can be detected with a camera.

The aim of this pilot study is to investigate the feasibility of camera based PPG for contactless HR monitoring in newborn infants in the NICU with ambient light. We did not use dedicated illumination as it might hinder the infant and would increase the experimental footprint in the NICU.

Section snippets

Study design

Infants were studied in the NICU in both the Children's Hospital of Orange County (CHOC), California, USA and in the Máxima Medical Center (MMC), The Netherlands. Institutional Review Board approval (CHOC#090768 and UCI#2009-7046, clinicaltrials.gov: NCT00989859, Philips Research-MMC #2010-075) and informed parental consent were obtained prior to measurements. Since the objective was to explore potential challenging conditions of the technique no exclusion criteria were defined and any infant

Results

A total of 19 infants with gestational ages ranging from 25 to 42 weeks, postnatal ages ranging from 3 days to 4 weeks, and weights from 470 to 3810 g were studied. Two were ventilated, one of which with HFOV. One infant had sepsis. No inotropics had been administered to the infants at the time of the measurement. Ambient light levels in the NICU varied from 20 to 180 lx. For comparison, typical office work floor light intensities range from 400 to 750 lx. During phototherapy to treat neonatal

Discussion

The most fundamental condition for camera based PPG to be feasible in the NICU is that the cardiovascular induced PPG signal has to be detectable in the infants at an arbitrary anatomical location. This means that non-cardiovascular related changes in skin optical reflectivity (such as caused by redistribution of venous blood, e.g. during sepsis or by repositioning a limb causing changes in hydrostatic pressure) have to be smaller than those induced by the cardio-vascular pulse wave. We could

Conflict of interest statement

The authors have no conflict of interest to declare.

References (20)

  • C. Takano et al.

    Heart rate measurement based on a time-lapse image

    Med Eng Phys

    (2007)
  • C.J. Upton et al.

    Episodic bradycardia in preterm infants

    Arch Dis Child

    (1992)
  • A. Ohlin et al.

    Clinical signs and CRP values associated with blood culture results in neonates evaluated for suspected sepsis

    Acta Paediatr

    (2010)
  • A.B.. Hertzman et al.

    Observations on the finger volume pulse recorded photo-electrically

    Am J Physiol

    (1937)
  • M.M. Baharestani

    An overview of neonatal and pediatric wound care knowledge and considerations

    Ostomy Wound Manage

    (2007)
  • F.S. Afsar

    Skin care for preterm and term neonates

    Clin Exp Dermatol

    (2009)
  • Kuller J. McManus

    Skin breakdown: risk factors, prevention, and treatment

    Newborn Infant Nurs Rev

    (2001)
  • K.J. Anand et al.

    Can adverse neonatal experiences alter brain development and subsequent behavior?

    Biol Neonate

    (2000)
  • W. Chen et al.

    Intelligent design for neonatal monitoring with wearable sensors

  • F.P. Wieringa et al.

    Contactless multiple wavelength photoplethysmographic imaging: a first step toward “SpO2 camera” technology

    Ann Biomed Eng

    (2005)
There are more references available in the full text version of this article.

Cited by (193)

  • Remote photoplethysmography for heart rate measurement: A review

    2024, Biomedical Signal Processing and Control
  • A non-contact oxygen saturation detection method based on dynamic spectrum

    2022, Infrared Physics and Technology
    Citation Excerpt :

    It not only changes the body temperature, but also affects the variability of the person's oxygen saturation [4], and the ability to achieve rapid non-contact measurements in congregate locations helps to avoid cross-contamination. Imaging photoplethysmography (iPPG), with its advantages of non-contact and large information content in clinical applications, has attracted many scholars to explore[5–7], including heart rate [8–11], oxygen saturation [12–14], respiratory rate, and blood pressure [15–17]. Wieringa et al. [18] first validated the feasibility of non-contact blood oxygen using a camera and a light source of three wavelengths measurement, but it was not well used due to limitations such as low SNR (signal-to-noise ratio).

View all citing articles on Scopus
1

Current affiliation: Department of Pediatrics, Rijnstate Hospital, Wagnerlaan 55, 6800 TA Arnhem, The Netherlands.

View full text