Table 3.

Studies for neonatal motion quantification using indirect sensing technology. The order is sorted by the aim of motion quantification, and year for each sensing technology.

Author (Refer)	Year	Subject size	Sensing technology	Product model	Placement	Aim of motion quantification	Method of motion quantification
Ossmy et al. [21]	2020	I = 1	RGB Camera	Not mentioned	Overhead camera view	Motion/position monitoring: Generate the moving path of the infant	Changes in feet location detected by convolutional pose machine
Zhao et al. [22]	2020	I = 1	RGB Camera	OV5642	On the ceiling of the incubator	Motion/position monitoring: Trigger alarm system to notify caregivers	Background subtraction
Peng et al. [23]	2022	I = 18	RGB/Thermal Camera	UI-3860LE-C-HQ/FLIR Lepton5.5	On top of the incubator or bed	Motion/position monitoring: Motion detection	Background subtraction, optical flow, ORB
Mazzarella et al. [24]	2020	I = 14	RGB Camera	10-camera Vicon motion capture system	In front of infants	Neuromotor pathology: Extract features from quantified motion signals to detect perinatal stroke and cerebral palsy	Changes in 3D coordinates of markers
Malik et al. [25]	2020	I = 2	RGB Camera	Not mentioned	Bedside	Neuromotor pathology: Magnify motion signal to detect tremors	Convolutional encoder-manipulator-decoder network
Wu et al. [26]	2021	I = 59	RGB/RGBD Camera	Kinect	On top of the bed	Neuromotor pathology: Extract features from identified 14-keypoint motion sequence to predict cerebral palsy	Changes in coordinates of 14 keypoints detected by PifPaf
Zamzmi et al. [27]	2016	I = 18	RGB Camera	Not mentioned	On top of the bed	Pain/discomfort detection: Extract a feature from a quantified motion signal to assess pain	Background subtraction
Sun et al. [28]	2019	P = 11	RGB Camera	uEye UI-222x	On the ceiling of the incubator	Pain/discomfort detection: Extract features from quantified motion signals to detect discomfort	Optical flow
Sun et al. [29]	2021	I = 24	RGB Camera	Xacti VPC-FH1BK	On top of the bed (only face view)	Pain/discomfort detection: Quantified motion signals combined with RGB images as inputs of neural network for discomfort detection	Optical flow
Ferrari et al. [30]	2010	I = 1	RGB Camera	Not mentioned	In front of the bed	Seizure detection: Identify periodicity of motion to detect clonic seizures	Background subtraction
Pisani et al. [31]	2014	I = 12	RGB Camera	Not mentioned	In front of bed	Seizure detection: Identify periodicity of motion to detect clonic seizures	Background subtraction
Martin et al. [32]	2022	I = 43	RGB Camera	Cadwell systems	On top of the bed	Seizure detection: Extract power from the quantified motion signal to detect seizure	Optical flow
Mestha et al. [33]	2014	I = 8	RGB Camera	Not mentioned	Bedside	Vitals: Reject unreliable period with motion artifact for pulse rate estimation.	Background subtraction
Rossol et al. [34]	2020	I = 2	RGB (YUV) Camera	Wansview	4–6 feet on top of the bed	Vitals: Extract respiration from the motion signal	Background subtraction
Lorato et al. [35]	2021	P = 5 I = 12	RGB/Thermal Camera	UI-2220SE/FLIR Lepton5.5	Bedside	Vitals: Reject unreliable period with motion artifact for respiration estimation	Background subtraction
Lyra et al. [36]	2022	I = 19	Thermography Camera	VarioCAM HD head 820 S	On top of the bed	Vitals: Reject unreliable period with motion artifact for respiration estimation	Displacement of the bounding box of the head detected by YOLOv4-Tiny
Long et al. [37]	2018	I = 5	IR Camera	Philips Avent uGrow baby monitor	In front of the bed	Others: Detect infant presence and in/out bed motion	3D recursive search
Chaichulee et al. [38]	2019	P = 15	RGB Camera	3-CCD JAI AT-200CL digital video camera (JAI A/S, Denmark)	Inside the incubator through a hole	Others: Quantified motion signals combined with image signals as inputs of neural networks for intervention detection	Optical flow
Andrea et al. [39]	2022	P = 201 I = 501	In-bore Camera	Not mentioned	Inside MRI	Others: Motivate MRI operators to adjust scan protocols accordingly	Framewise displacement
Lee et al. [40]	2020	P = 16 I = 18	Radar	IR-UWB Radar	35 cm orthogonal away from the chest	Vitals: Reject unreliable period with motion artifact for cardiorespiratory monitoring	Power differences
Beltrao et al. [41]	2022	P = 12	Radar	Radar in 24-GHZ ISM band	45–50 cm orthogonal away from the chest	Vitals: Identify and mitigate motion artifacts for respiration estimation	Nonnegative matrix factorization
Joshi et al. [42]	2018	P = 10	Mattress/BSG	electromechanical film sensor (EMFi, Emfitt, Kuopio, Finland)	On top of the regular mattress and covered by the bedsheet	Motion/position monitoring: Gross motion detection Motion Monitoring	Signal instability index from BSG
Aziz et al. [43]	2020	I = 5	Mattress/BSG	LX100:100.100.05 (XSensor Technology Corp)	Between bedding and infant	Motion/position monitoring: Gross motion detection	Displacement of the center of the pressure
Ranta et al. [44]	2021	I = 43	Mattress/BSG	Electromechanical ferroelectric sensor (L-4060SLC, Emfit, Finland)	Not mentioned	Sleep: Extract motion-related features for sleep stage classification	Smoothed root mean square value of BSG
Williamson et al. [45]	2013	P = 6	PPG reusing	Not appliable	Not appliable	Apnea detection: Extract motion-related features for apnea prediction	Low-frequency power in PPG
Zuzarte et al. [46]	2021	P = 10	PPG reusing	Not appliable	Not appliable	Apnea detection: Extract motion-related features for apnea prediction	Low-frequency power in PPG
Zuzarte et al. [3]	2019	P = 18	PPG reusing	Not appliable	Not appliable	Motion/position monitoring: Gross motion detection	Low-frequency power in PPG
Peng et al. [47]	2021	I = 15	ECG, PPG, CI reusing	Not appliable	Not appliable	Motion/position Monitoring: Gross motion detection	Signal instability index and/or Low-frequency power
Cabrera-Quiros et al. [48]	2021	P = 64	ECG reusing	Not appliable	Not appliable	Sepsis detection: Extract motion-related features for sepsis prediction	Signal instability index from ECG
Peng et al. [49]	2022	P = 127	ECG, CI reusing	Not appliable	Not appliable	Sepsis detection: Extract motion-related features for sepsis prediction	Signal instability index and/or Low-frequency power

P – Preterm infants; I – Infants; BSG -- ballistography; PPG -- photoplethysmogram; ECG -- electrocardiogram; CI – chest impedance.