Detection of respiratory activity in newborn infants using a noncontact vision-based monitor.

OBJECTIVE: To examine the effectiveness of a noncontact vision-based infrared respiratory monitor (IRM) in the detection of authentic respiratory motion in newborn infants. STUDY DESIGN: Observational study in a neonatal intensive care unit. METHODS: Eligible infants lay supine with torso exposed under the IRM's infrared depth-map camera and torso images were recorded at 30 frames/s. Respiratory motion waveforms were subsequently derived from upper (IRMupper ) and lower (IRMlower ) torso region images and compared with contemporaneous impedance pneumography (IP) and capsule pneumography (CP). Waveforms, in 15 s investigative epochs, were scanned with an 8 s sliding window for authentic respiratory waveform (spectral purity index [SPI] ≥ 0.75, minimum five complete breaths). Maximum SPI and frequency of occurrence of authentic respiratory waveform in 15 s epochs were compared between monitoring modalities in pooled and per patient data (Friedman ANOVA). RESULTS: Recordings comprised 532 min of images from 35 infants, yielding 2131 investigative epochs, with authentic respiratory motion detected in all infants. For CP, IP, IRMupper , and IRMlower , the proportion of epochs containing authentic respiratory motion in pooled data were 65%, 50%, 36%, and 48%, with median SPImax of 0.79, 0.75, 0.70, and 0.74, respectively. Per-patient average SPImax was 0.79, 0.75, 0.69, and 0.74 for CP, IP, IRMupper , and IRMlower with proportion of authentic respiratory motion being 64%, 50%, 29%, and 49%, respectively. CONCLUSION: An IRM focused on the lower torso detected authentic respiratory motion with comparable performance to IP in newborn infants in intensive care and deserves further investigation.

Automated control of oxygen titration in preterm infants on non-invasive respiratory support.

OBJECTIVE: To evaluate the performance of a rapidly responsive adaptive algorithm (VDL1.1) for automated oxygen control in preterm infants with respiratory insufficiency. DESIGN: Interventional cross-over study of a 24-hour period of automated oxygen control compared with aggregated data from two flanking periods of manual control (12 hours each). SETTING: Neonatal intensive care unit. PARTICIPANTS: Preterm infants receiving non-invasive respiratory support and supplemental oxygen; median birth gestation 27 weeks (IQR 26-28) and postnatal age 17 (12-23) days. INTERVENTION: Automated oxygen titration with the VDL1.1 algorithm, with the incoming SpO2 signal derived from a standard oximetry probe, and the computed inspired oxygen concentration (FiO2) adjustments actuated by a motorised blender. The desired SpO2 range was 90%-94%, with bedside clinicians able to make corrective manual FiO2 adjustments at all times. MAIN OUTCOME MEASURES: Target range (TR) time (SpO2 90%-94% or 90%-100% if in air), periods of SpO2 deviation, number of manual FiO2 adjustments and oxygen requirement were compared between automated and manual control periods. RESULTS: In 60 cross-over studies in 35 infants, automated oxygen titration resulted in greater TR time (manual 58 (51-64)% vs automated 81 (72-85)%, p<0.001), less time at both extremes of oxygenation and considerably fewer prolonged hypoxaemic and hyperoxaemic episodes. The algorithm functioned effectively in every infant. Manual FiO2 adjustments were infrequent during automated control (0.11 adjustments/hour), and oxygen requirements were similar (manual 28 (25-32)% and automated 26 (24-32)%, p=0.13). CONCLUSION: The VDL1.1 algorithm was safe and effective in SpO2 targeting in preterm infants on non-invasive respiratory support. TRIAL REGISTRATION NUMBER: ACTRN12616000300471.

Automatic Torso Detection in Images of Preterm Infants.

Imaging systems have applications in patient respiratory monitoring but with limited application in neonatal intensive care units (NICU). In this paper we propose an algorithm to automatically detect the torso in an image of a preterm infant during non-invasive respiratory monitoring. The algorithm uses normalised cut to segment each image into clusters, followed by two fuzzy inference systems to detect the nappy and torso. Our dataset comprised overhead images of 16 preterm infants in a NICU, with uncontrolled illumination, and encompassing variations in poses, presence of medical equipment and clutter in the background. The algorithm successfully identified the torso region for 15 of the 16 images, with a high agreement between the detected torso and the torso identified by clinical experts.

Eye-screen distance monitoring for computer use.

The extended period many people now spend looking at computer screens is thought to affect eyesight over the long term. In this paper we are concerned with developing and initial evaluation of a wireless camera-based tracking system providing quantitative assessment of computer screen interaction. The system utilizes a stereo camera system and wireless XBee based infrared markers and enables unobtrusive monitoring. Preliminary results indicate that the system is an excellent method of monitoring eye-screen distance. This type of system will enable future studies of eye-screen distance for computer users.

A single camera stereographic imaging system for tracking upper limb motions.

A single camera stereographic imaging system was designed, built and tested for the intended application of tracking upper limb motion. The system consisted of a camera, quad-mirror system and image acquisition and processing software. This paper describes the design methods used and the testing of the system. Design methods included geometry optimisation and testing consisted of proof-of-concept trials. Preliminary results show that the concept is sound.