Camera-Based On-Line Short Cessation of Breathing Detection

Self Cite

Aiming at continuous unobtrusive respiration monitoring, motion robustness is paramount. However, some types of motion can completely hide the respiration information and the detection of these events is required to avoid incorrect rate estimations. Therefore, this work proposes a motion detector optimized to specifically detect severe motion of infants combined with a respiration rate detection strategy based on automatic pixels selection, which proved to be robust to motion of the infants involving head and limbs. A dataset including both thermal and RGB (Red Green Blue) videos was used amounting to a total of 43 h acquired on 17 infants. The method was successfully applied to both RGB and thermal videos and compared to the chest impedance signal. The Mean Absolute Error (MAE) in segments where some motion is present was 1.16 and 1.97 breaths/min higher than the MAE in the ideal moments where the infants were still for testing and validation set, respectively. Overall, the average MAE on the testing and validation set are 3.31 breaths/min and 5.36 breaths/min, using 64.00% and 69.65% of the included video segments (segments containing events such as interventions were excluded based on a manual annotation), respectively. Moreover, we highlight challenges that need to be overcome for continuous camera-based respiration monitoring. The method can be applied to different camera modalities, does not require skin visibility, and is robust to some motion of the infants.

Section: Discussionmentioning

confidence: 99%

Towards Continuous Camera-Based Respiration Monitoring in Infants

Lorato

Stuijk

Meftah

et al. 2021

Self Cite

“…Using the chosen pixels, a RF signal and the flow-based Respiration Rate (RR) were obtained. Moreover, OAs were simulated in the 87 video segments and our previously proposed method for the detection of COBs [ 35 ] was used. This allows comparing the detectability of the OAs in the different respiration signals we obtain from the thermal videos (i.e., MR signal and RF signal ).…”

Section: Methodsmentioning

confidence: 99%

“…We adjusted our previously published COB-detector [ 35 ] to evaluate the detectability of OAs, which were simulated as indicated in Section 2.2.4 . The COB-detector assumes that COBs can be detected by monitoring sudden amplitude changes and it is based on the comparison of a short-term standard deviation and a long-term standard deviation .…”

Section: Methodsmentioning

confidence: 99%

“…The only adaptations applied to our previous published implementation concern the length of the windows for the calculation of the two standard deviations. The duration of these windows was chosen in [ 35 ] based on the targeted COB. In particular, the window for the calculation of the short-term standard deviation should be close to the minimum COB duration, an apnea of 10 s. Although, the window for the long-term standard deviation , which is calculated as median of the short-term standard deviations , must be higher than the COB duration.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Automatic Separation of Respiratory Flow from Motion in Thermal Videos for Infant Apnea Detection

Lorato

Stuijk

Meftah

et al. 2021

Self Cite

Both Respiratory Flow (RF) and Respiratory Motion (RM) are visible in thermal recordings of infants. Monitoring these two signals usually requires landmark detection for the selection of a region of interest. Other approaches combine respiratory signals coming from both RF and RM, obtaining a Mixed Respiratory (MR) signal. The detection and classification of apneas, particularly common in preterm infants with low birth weight, would benefit from monitoring both RF and RM, or MR, signals. Therefore, we propose in this work an automatic RF pixel detector not based on facial/body landmarks. The method is based on the property of RF pixels in thermal videos, which are in areas with a smooth circular gradient. We defined 5 features combined with the use of a bank of Gabor filters that together allow selection of the RF pixels. The algorithm was tested on thermal recordings of 9 infants amounting to a total of 132 minutes acquired in a neonatal ward. On average the percentage of correctly identified RF pixels was 84%. Obstructive Apneas (OAs) were simulated as a proof of concept to prove the advantage in monitoring the RF signal compared to the MR signal. The sensitivity in the simulated OA detection improved for the RF signal reaching 73% against the 23% of the MR signal. Overall, the method yielded promising results, although the positioning and number of cameras used could be further optimized for optimal RF visibility.

“…There are several non-contact monitoring methods based on visual detection which have been tested on babies, such as thermal imaging [ 1 , 2 , 3 , 4 , 5 ] and Red Green Blue (RGB) cameras [ 6 , 7 , 8 , 9 ]. The main drawback of these methods is that the face needs to be visible and that RGB imaging is not independent of lighting conditions.…”

Section: Introductionmentioning

confidence: 99%

Automated Non-Contact Respiratory Rate Monitoring of Neonates Based on Synchronous Evaluation of a 3D Time-of-Flight Camera and a Microwave Interferometric Radar Sensor

Gleichauf

Herrmann

Hennemann

et al. 2021

This paper introduces an automatic non-contact monitoring method based on the synchronous evaluation of a 3D time-of-flight (ToF) camera and a microwave interferometric radar sensor for measuring the respiratory rate of neonates. The current monitoring on the Neonatal Intensive Care Unit (NICU) has several issues which can cause pressure marks, skin irritations and eczema. To minimize these risks, a non-contact system made up of a 3D time-of-flight camera and a microwave interferometric radar sensor is presented. The 3D time-of-flight camera delivers 3D point clouds which can be used to calculate the change in distance of the moving chest and from it the respiratory rate. The disadvantage of the ToF camera is that the heartbeat cannot be determined. The microwave interferometric radar sensor determines the change in displacement caused by the respiration and is even capable of measuring the small superimposed movements due to the heartbeat. The radar sensor is very sensitive towards movement artifacts due to, e.g., the baby moving its arms. To allow a robust vital parameter detection the data of both sensors was evaluated synchronously. In this publication, we focus on the first step: determining the respiratory rate. After all processing steps, the respiratory rate determined by the radar sensor was compared to the value received from the 3D time-of-flight camera. The method was validated against our gold standard: a self-developed neonatal simulation system which can simulate different breathing patterns. In this paper, we show that we are the first to determine the respiratory rate by evaluating the data of an interferometric microwave radar sensor and a ToF camera synchronously. Our system delivers very precise breaths per minute (BPM) values within the norm range of 20–60 BPM with a maximum difference of 3 BPM (for the ToF camera itself at 30 BPM in normal mode). Especially in lower respiratory rate regions, i.e., 5 and 10 BPM, the synchronous evaluation is required to compensate the drawbacks of the ToF camera. In the norm range, the ToF camera performs slightly better than the radar sensor.