This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
1
Wireless Monitoring System for Oral-Feeding Evaluation of Preterm Infants Chang-Ting Chen, Yu-Lin Wang, Chen-An Wang, Mei-Ju Ko, Wai-Chi Fang, Fellow, IEEE, and Bor-Shyh Lin, Senior Member, IEEE
Abstract—The oral feeding disorder is one of the important indicators for the high risk group of neurodevelopment delay. The procedure of oral feeding requires the coordination of sucking, swallowing, and breathing activities, and it is the most complex sensorimotor process for newborn infants. Premature infants often uneasily complete the procedure of oral feeding. However, the evaluation of the oral feeding disorders and severity are usually dependent on the subjective clinical experience of the physician. Monitoring the sucking-swallowing-breathing activities directly is difficult for preterm infants. In this study, a wireless monitoring system for oral-feeding evaluation of full term and preterm infants was proposed to objectively and quantitatively evaluate the coordination of suck-swallow-respiration function during oral feeding. Moreover, the ratios of the swallowing and breathing event numbers to the sucking event number were defined to evaluate the coordination of suck-swallow-respiration function during oral feeding. Finally, the system performance was validated and the coordination of suck-swallow-respiration function for full term and preterm infants during oral feeding was also investigated. Index Terms—Oral feeding disorder, premature infant, suck-swallow-respiration function, wireless monitoring system.
I. INTRODUCTION
P
REMATURE infants, that their gestation week is less than 37 week, usually have immature organs, and the immature organs always bring some risk factors. The oral feeding disorder is one of the common disease for premature infants. The procedure of oral feeding requires the precise coordination between sucking, swallowing and breathing activities. The premature infants may unsuccessfully complete the procedure
Manuscript received January 08, 2015; revised May 14, 2015; accepted May 23, 2015. The authors are greatly indebted to the Ministry of Science and Technology, Taiwan for the support of the research through contracts in 103-2221-E009-035-MY2. This paper was recommended by Associate Editor S. Sonkusale. C.-T. Chen and C.-A. Wang are with the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu 300, Taiwan. Y.-L. Wang is with the Department of Rehabilitation, Chi Mei Medical Center, Tainan 710, Taiwan, and also with the Department of Sports Management, Chia Nan University of Pharmacy and Science, Tainan City 717, Taiwan. M.-J. Ko is with the Department of Rehabilitation, Chi Mei Medical Center, Tainan 710, Taiwan. W.-C. Fang is with the Department of Electronics Engineering, National Chiao Tung University, Hsinchu 300, Taiwan. B.-S. Lin is with the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu 300, Taiwan, and also with the Department of Medical Research, Chi-Mei Medical Center, Tainan 710, Taiwan (e-mail: [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TBCAS.2015.2438031
of oral feeding, and this affects the absorption of nutrients and may result in bradycardia and apnea [1]. Many premature infants have to receive oral and non-oral sensorimotor interventions until these infants are able to independently complete nutritive sucking [2]. For preterm infants, oral motor intervention is a sensory stimulation of the lips, jaw, tongue, soft palate, pharynx, larynx and respiratory muscles pre-oral feeding, and assists to gain the rhythm of sucking-swallowing with monitoring stress signs (such as cyanosis, tachypnea, apnea or bradycardia) during oral feeding. The previous research has proved that the early oral motor intervention can shorten the transition time from gavage feeding to full oral feeding and improve oral feeding efficiency [3]. Non-oral sensorimotor intervention is a kind of tactile/ kinesthetic stimulations consisted of body stroking and passive movements of the limbs for three, 15-minute periods per day. The combined oral and non-oral sensorimotor intervention has an additive/synergistic effect for the independent complete nutritive sucking [4]. Preterm infants with heart or lung problems may need to breathe increased amounts of oxygen by the oxygen hood or nasal cannula to get normal levels of oxygen in their blood during oral feeding. Several previous studies further investigated the behavior of oral feeding for premature infants. Nutritive sucking is a variable process, and can be separated into the intermittent, continuous and pause phases during oral feeding from their sucking, swallowing, and breathing activities. During the first six months after birth, infants obtained their primary foods through nutritive sucking. By nutritive sucking, infants can be fed with lower energy requirement [5], [6]. However, the swallowing activity has to use laryngeal muscles, pharyngeal and oral simultaneously within a few milliseconds, and it is a difficult action for premature infants. Moreover, in order to prevent choking, the airways will be closed with the swallowing reflex motion [7]. Goldfield et al. indicated that the swallowing activity is not a random distribution during oral feeding procedure [8]. The event of swallowing activity happens before the next sucking event and between breathing out and breathing in. Infants cannot maintain ventilation, when the coordination of suck-swallow-respiration function collapses. Muller et al. investigated the weaning process of infants by using mechanical ventilators [9], and they indicated the infants’ diaphragmatic fatigue is highly related to their breathing activities. Fucile et al. investigated the suck-swallow-respiration function of the preterm infants during oral feeding, and they indicated that applying the oral sensory stimulation to infants can improve the ability of oral feeding [10]. Therefore, the parameters related to the coordination between sucking,
1932-4545 © 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 2
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
swallowing, and breathing activities are the key indicators for the oral feeding ability of premature infants. Up to now, the oral feeding ability is evaluated and diagnosed subjectively by medical staffs. Thus, developing the monitoring system to objectively and quantitatively evaluate the suck-swallow-respiration function of premature infants is important for the diagnosis of oral feeding disorders. Several methods have been proposed to monitor the swallowing or breathing activities. The accelerometer-based or sEMG-based approaches have been proposed to monitor the swallowing activity [11], [12]. Cheng et al. proposed a textile capacitive sensor to measure the capacitance variation of the human body and monitor swallowing activity [11]. Sejdic et al. used a dual-axis accelerometer to monitor the movement of the throat, and applied it to detect the state of dysphagic patients [12]. The approaches of using the acceleration [13], the respiratory belt [14], or microphone [15] have also been used to monitor the breathing activity in previous studies. However, premature infants are fragile and their muscle activities are unapparent. Therefore, the above methods are not suitable to monitor the swallowing and breathing activities of premature infants. Additionally, there is still lack of medical instruments for monitoring the sucking activity of infants directly during oral feeding due to that the electrical sensor is easily affected and destroyed by milk. In this study, a wireless monitoring system for oral-feeding evaluation of preterm infants was proposed to quantitatively evaluate the coordination of suck-swallow-respiration function during oral feeding. Here, a novel sensing device was designed and implemented to monitor the sucking pressure of infants during oral feeding. The raw surface electromyography (sEMG), measured from the chest of infants, and the swallowing sound were used to detect the events of breathing and swallowing activities respectively. Moreover, the ratios of the swallowing and breathing event numbers to the sucking event number were defined and used to evaluate the coordination of suck-swallow-respiration function during oral feeding. Finally, the system performance was validated and the coordination of suck-swallow-respiration function for premature infants and full-term infants during oral feeding was also investigated. II. SYSTEM ARCHITECTURE AND DESIGN The basic framework and photograph of the proposed wireless monitoring system for oral-feeding evaluation are shown in Fig. 1(a) and (b) respectively. The proposed system mainly contains a sensing device for monitoring sucking pressure, a wireless multi-parameter acquisition module, and a back-end system platform. Here, the nipple of the sensing device for sucking pressure was placed in the mouth of the infant to measure the sucking pressure under oral feeding, and convert the sucking pressure into electrical signal. The swallowing sound and the raw sEMG related to the breathing activity were acquired by a mini microphone and the Neonatal Electrodes (Kendall™ 1041PTS, Covidien, Ireland) respectively. Here, the mini microphone was placed on the neck (at the lower border of the cricoid cartilage), and a pair of surface electrodes were placed at the merging of the sixth and seventh intercostal space along the anterior axillary line with grounding electrode
Fig. 1. (a) Basic framework and (b) photograph of proposed wireless monitoring system.
at the xiphoid. Subsequently, these electrical signals of sucking pressure, swallowing sound and raw breathing sEMG were amplified, filtered, and digitized by the wireless multi-parameter acquisition module, and then were transmitted to the back-end system platform wirelessly via Bluetooth. Finally, the multi-parameter monitoring program built in the back-end system platform received these bio-signals and detected the events of sucking, swallowing, and breathing activities. The ratio of the swallowing event number to the sucking event number and the ratio of the breathing event number to the sucking event number were also calculated by the multi-parameter monitoring program to evaluate the ability of oral feeding for infants. A. Sensing Device for Monitoring Sucking Pressure Fig. 2 shows the basic structure and photograph of the sensing device for sucking pressure. It mainly consists of an electrical pressure sensor, a rubber tube and a polypropylene bottle. In this study, intra-nipple (teat) pressure was measured with the pressure sensor (SSC-SNBN400MD-AA3, Honey Well, U. S.), which was connected to plastic tubing positioned to lie within the feeding nipple teat. Its measuring range is within , and the range of operation temperature is from Celsius to 85 degrees Celsius. The pressure sensor contains two pressure inputs, and the pressure difference between the two pressure inputs will be convert into a voltage signal. However, it is difficult to measure the sucking pressure directly during oral feeding, because the milk easily
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. CHEN et al.: WIRELESS MONITORING SYSTEM FOR ORAL-FEEDING EVALUATION OF PRETERM INFANTS
Fig. 2. Basic structure and photograph of proposed sensing device for monitoring sucking pressure.
3
unit. Here, the front-end amplifier circuits consist of pre-amplifiers and band-pass filters. The operational amplifier (AD8609, Analog Devices. U.S) was used in the front-end amplifier. The gains of the pre-amplifiers and the frequency band of the band-pass filters were designed according to the characteristics of different bio-signals. The gains of the pre-amplifiers for swallowing sound and breathing sEMG were set to 900 times and 4000 times respectively. Here, the 2-order Butterworth band-pass filters were implemented. The frequency band of the band-pass filter for sucking pressure and breathing sEMG were set to 0.5 Hz – 500 Hz and 0.5 Hz – 500 Hz respectively, and the frequency band of the high-pass filter for swallowing sound was set to . After amplifying and filtering these bio-signals, a 12-bit analog-to-digital converter (ADC) built in the microprocessor was used to digitize these bio-signals. The sampling rates for sucking pressure, swallowing sound, and breathing sEMG were set to 1000 Hz simultaneously. In this study, the microprocessor (MSP430, Texas Instruments, U. S.) was used to control the analog-to-digital converter, package the data of these bio-signals, and transmit them to the wireless transmission circuit. The design of the wireless transmission circuit was based on a Bluetooth module, supporting the Bluetooth v2.0 + EDR specification, with a printed circuit board-based antenna. The power management unit provided the stable 3 V power supply and the function of charging the battery. Here, a commercial 250 mAh Li-ion battery was used in this module, and this module can be operated continuously for 9 hours. C. Back-End System Platform
Fig. 3. Block diagram and photograph of wireless multi-parameter acquisition module.
affects and destroys the electrical pressure sensor. In order to overcome the above issue, the electrical pressure sensor was placed at the outside of the polypropylene bottle, and the rubber tube was used to connect with the electrical pressure sensor and the polypropylene bottle. Here, one terminal of the rubber tube was connected with the input of the pressure sensor, and the other terminal was placed at the nipple inside of the polypropylene bottle. Based on the connected pipe, the intra-nipple (teat) pressure change during sucking motions can be transmitted to the input of the pressure sensor. By comparing with the general atmospheric pressure, the sucking pressure of the nipple during oral feeding was measured indirectly by the pressure sensor to avoid the influence of milk during oral feeding. B. Wireless Multi-Parameter Acquisition Module Fig. 3 illustrates the block diagram and photograph of the wireless multi-parameter acquisition module. The size of the proposed module is about 8.5 5 1.8 cm3. It mainly contains several parts: front-end amplifier circuits, a microprocessor unit, a wireless transmission circuit, and a power management
The design of the back-end system platform in this system was based on a commercial laptop, and Windows 8.1 was used as the operation system of the system platform. A multi-parameter monitoring program built in the back-end system platform was developed by Microsoft C#, and it can provide the basic functions of receiving, displaying, and storing raw data, detecting the event number of sucking, swallowing, and breathing activities, and calculating the ratios of the event numbers of swallowing and breathing activity to that of sucking activity. III. SYSTEM SOFTWARE DESIGN The operation procedure and the screenshot of the multi-parameter monitoring program are shown in Fig. 4(a) and (b) respectively. The multi-parameter monitoring program mainly consists of three parts: GUI, BUFFER, and THREAD. Here, BUFFER is a container used to store data received from the wireless multi-parameter acquisition module. The part of GUI provided the ability to precisely control the location and display of the GUI elements, such as the form, panel, et al. The part of THREAD denotes the execution thread, and it contains the threads of BT API, REC, and ANALYSIS. Here, the thread of BT API and REC were designed to connect with the wireless multi-parameter acquisition module via Bluetooth, and receive the raw data and store them in BUFFER respectively. The design of the ANALYSIS thread was based on the multi-parameter analysis algorithm, which was designed to detect the event number of sucking, swallowing, and breathing activities, and the ratios of the event numbers of swallowing and breathing
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 4
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
Fig. 4. (a) Operation procedure and (b) screenshot of multi-parameter monitoring program.
activity to that of sucking activity to estimate the coordination of suck-swallow-respiration function. The procedure of the multi-parameter analysis algorithm is illustrated in Fig. 5. First, the raw signals of swallowing sound and breathing sEMG was pre-processed by high-pass filters to remove the low frequency noise and the interference of electrocardiogram (ECG). Next, the technique of fractal dimension (FD) [16] was applied to estimate the variation of signal complexity for the swallowing sound and breathing sEMG. The variation of the signal complexity was used as the event index of the swallowing and breathing activities. The fractal dimension value can be calculated by
(1) where denotes the total length of the curve, is the largest distance between the first point and any other point on the curve. Here, is defined as the number of steps in the curve, and can be calculated by (2) is the average distance between each consecutive where points. Here, a two second moving window with the overlap of one second was used to calculate the fractal dimension value. When the activity event occurred, the signal complexity of the biosignal increased to result in the increase of the fractal dimension value. Therefore, the events of the swallowing
Fig. 5. Flowchart of multi-parameter analysis algorithm.
and breathing activities reflected on the change of the fractal dimension value indirectly. Next, the positive peaks of the sucking pressure, and the fractal dimension values of the swallowing sound and breathing sEMG were detected by using the first derivative (FDI) approach [17], and these positive
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. CHEN et al.: WIRELESS MONITORING SYSTEM FOR ORAL-FEEDING EVALUATION OF PRETERM INFANTS
5
TABLE I PERFORMANCE OF DETECTING SUCKING EVENTS BY PROPOSED ALGORITHM
PERFORMANCE
TABLE II DETECTING SWALLOWING EVENTS PROPOSED ALGORITHM
OF
BY
TABLE III PERFORMANCE OF DETECTING BREATHING EVENTS BY PROPOSED ALGORITHM
Fig. 6. Number of sucking, swallowing, and breathing events corresponding to different infant ages.
The sensitivity and PPV for detecting the event of the swallowing activity were 92.89% and 91.31%. The sensitivity and PPV for detecting the event of the breathing activity were 96.51% and 87.18%. peaks were viewed as the events of sucking, swallowing, and breathing activities. Finally, the event numbers of sucking, swallowing, and breathing activities per one minute were used to calculate the ratio and to estimate the coordination of suck-swallow-respiration function during oral feeding. Here, the parameters and denote the ratio of the swallowing event number to the sucking event number and the ratio of the breathing event number to the sucking event number respectively. IV. RESULTS A. Performance of Multi-Parameter Analysis Algorithm In this section, the performance of the proposed multi-parameter analysis algorithm was investigated. Here, several types of binary classification test parameters for validating the algorithm performance were first defined, and listed as follows: true positive (TP) means the activity event is correctly detected as an event, false positive (FP) means that no activity event is wrongly detected as an event, false negative (FN) means that the activity event is wrongly detected as nothing, and true negative (TN) means that no activity event is correctly detected as nothing. Tables I–III show the performance of detecting the events of the sucking, swallowing, and breathing activities by using the proposed algorithm respectively. For detecting the event of the sucking activity, the sensitivity and positive predictive value (PPV) were 96.7% and 94.25%. Here, the positive predictive value denotes the proportions of positive results in statistics that are true positive results and sensitivity denotes the proportion of actual positives which are correctly identified as positives, and they can be given by (3) (4)
B. Coordination Evaluation of Suck-Swallow-Respiration Function In this section, the coordination of suck-swallow-respiration function for preterm infants and full-term infants were evaluated. There were 22 for full-term ( ) Asian infants, 26 for 34 weeks, 9 for 35 weeks and 5 for 36 weeks of postmenstrual age premature Asian infants that attended this experiment. This experiment has been approved by the Institutional Review Board (IRB), Chi-Mei medical center, Taiwan, and all participants provided an informed consent. Fig. 6 shows the event number of sucking, swallowing, and breathing activities corresponding to different infant ages. The event numbers of the sucking and swallowing activities increased with the increase of the infant age. However, the event number of the breathing activity decreased with the increase of the infant age. For the full-term infants, the ratio of the sucking, swallowing, and breathing activities was about 1:1:1. Fig. 7 shows the ratios and for the infants with different ages. The analysis of variance (ANOVA) method was used for the analysis, and the difference significance was defined as in this study. Both of the ratio and for preterm infants were significantly different from that of full-term infants. The ratio was close to 1, and its standard deviation decreased with the increase of the infant age. The ratio of decreased obviously and became more stable with the increase of the infant age. For the full-term infant, both of the ratio and became close to 1, and were relatively stable. V. DISCUSSIONS Several monitoring systems have been developed in previous studies to detect swallowing and breathing activities. Table IV shows the system comparison between the proposed system and other monitoring systems. Different from the neck of the adult, the neck of the premature infant is very short, and the muscle movement of the infant is unapparent when
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 6
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
TABLE IV SYSTEM COMPARISON BETWEEN PROPOSED SYSTEM AND OTHER MONITORING SYSTEMS
Fig. 7. Ratios of RSS and RBS corresponding to different infant ages. Here, the sign * denotes the significant difference.
swallowing. Attaching the surface electrodes to the infant’s neck is difficult and inconvenient, and the swallowing sEMG is also easily affected by other biological potentials. By replacing the accelerometer-based or sEMG-based approaches used in the previous studies, the proposed system used the microphone to successfully monitor the swallowing activity. Moreover, by replacing the approaches of using the respiratory belt and the acceleration, the proposed system used sEMG to successfully monitor the breathing activity, and improve the issue of the chest movement of infants is very small when breathing. The sensing device for monitoring sucking pressure was also designed to successfully monitor the intra-nipple (teat) pressure. By the special mechanical design of the sensing device, it successfully avoided the influence of milk on the electrical pressure sensor [18], [19]. Several algorithms have also been developed to detect the breathing and swallowing activities in previous studies. In 2009, Sazonov et al. proposed the ECG-derived respiration technique which was based on the technique of wavelet decomposition to extract the information of breathing activity
from one lead ECG signal [20], and its mean error rate was about . In 2009, Sazonov et al. used the approach of support vector machine with mel-scale Fourier spectrum and wavelet packet decomposition techniques to detect swallowing activity, and its accuracy of detecting swallowing activity was about 84.7% [21]. In 2010, Damouras et al. proposed the swallow detection algorithm based on the quadratic variation technique to detect swallowing activity from the dual-axis accelerometry signal, and its accuracy was more than 90% [22]. By using the technique of fractal dimension, the events of swallowing and breathing activities were exactly reflected on the change of their fractal dimension values. Most of the sucking, swallowing, and breathing events were effectively detected by the proposed algorithm. From the experimental results, the state of false positive usually occurred when a longer and continuous activity event was detected as a number of separate activity events. The state of false negative usually occurred when a short or unapparent activity event happened. In this study, the ratios of and were also defined to evaluate the coordination of suck-swallow-respiration function for infants. Both of the ratio and became smaller and more stable with the increase of the infant age. From the concept of cross-systems interactions, in order to ensure the infant feeding safely, central pattern generators (CPGs) integrate and coordinate the motor neurons of swallowing, sucking, and breathing [18]. With the gradual maturation of the infant, the sucking and swallowing abilities of the infant become more proficient, and this also results in that the sucking and swallowing events become more rapid and coordinated. In general, the ratio of the sucking events to the swallowing event for a healthy full-term infant is about 1:1 or 2:1 [23]. Therefore, the ratio of was close to 1, and became more stable when the infant age increased. The rate of respiratory activity is also an important indicator to evaluate the ability of oral feeding for infants. During the continuous sucking phase of oral feeding, the rate of respiratory activity for the preterm infants is about 26–31 events per minute, and that for full-term infants is about 30–35
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. CHEN et al.: WIRELESS MONITORING SYSTEM FOR ORAL-FEEDING EVALUATION OF PRETERM INFANTS
events per minute [24], [25]. The preterm infant needs more prolonged breath in the alternating periods of sucking, swallowing, and breathing activities. Because the prolonged breath pauses in the alternating periods or blocking the nipple with the tongue, the preterm infants have poor and irregular coordination of sucking, swallowing, and breathing activities during oral feeding. When the preterm infants matures, the ratio of sucking events to breathing events will also trend to 1:1 [26], [27]. Therefore, the ratio of also became close to 1 and more stable when the infant age increased.
VI. CONCLUSIONS In this study, a wireless monitoring system was proposed to continuously and non-invasively monitor the sucking-swallowing-breathing activities during oral feeding and evaluate the ability of oral feeding for premature infants. The sensing device for sucking pressure was also successfully applied in monitoring the sucking pressure during oral feeding to avoid the influence of milk. From the experimental results, the events of sucking-swallowing-breathing activities were detected effectively by the proposed multi-parameter analysis algorithm. Moreover, the ratios of the swallowing and breathing event number to the sucking event number were also defined to evaluate the coordination of suck-swallow-respiration function for infants. The ratio of for premature infants was significantly larger than that of full-term infants. Both of the ratio and became stable with the increase of the infant age. For full-term infants, the ratio of sucking, swallowing, and breathing was close to 1:1:1.
REFERENCES [1] C. Slocum, M. Arko, J. Di Fiore, R. J. Martin, and A. M. Hibbs, “Apnea, bradycardia and desaturation in preterm infants before and after feeding,” J. Perinatol., vol. 29, pp. 209–212, 2009. [2] B. Medoff-Cooper and W. Ray, “Neonatal sucking behaviors,” J. Nursing Scholar., vol. 27, pp. 195–200, 1995. [3] J. Arvedson, H. Clark, C. Lazarus, T. Schooling, and T. Frymark, “Evidence-based systematic review: Effects of oral motor interventions on feeding and swallowing in preterm infants,” Amer. J. Speech-Language Pathol., vol. 19, no. 4, pp. 321–340, 2010. [4] S. Fucile, E. G. Gisel, D. H. McFarland, and C. Lau, “Oral and non-oral sensorimotor interventions enhance feeding performance in preterm infants,” Develop. Med. Child Neurol., vol. 53, pp. 829–835, 2011. [5] J. Koenig, A. Davies, and B. Thach, “Coordination of breathing, sucking, and swallowing during bottle feedings in human infants,” J. Appl. Phys., vol. 69, pp. 1623–1629, 1990. [6] B. Medoff-Cooper, W. Bilker, and J. M. Kaplan, “Sucking patterns and behavioral state in 1-and 2-day-old full-term infants,” J. Obstetric, Gynecol., Neonatal Nursing, vol. 39, pp. 519–524, 2010. [7] J. F. Bosma, “Postnatal ontogeny of performances of the pharynx, larynx, and mouth,” Amer. Rev. Respiratory Disease, vol. 131, pp. 10–15, 1985. [8] E. C. Goldfield, C. Buonomo, K. Fletcher, J. Perez, S. Margetts, A. Hansen, V. Smith, S. Ringer, M. J. Richardson, and P. H. Wolff, “Premature infant swallowing: Patterns of tongue-soft palate coordination based upon videofluoroscopy,” Infant Behav. Develop., vol. 33, pp. 209–218, 2010. [9] N. Muller, G. Volgyesi, M. H. Bryan, and A. C. Bryan, “The consequences of diaphragmatic muscle fatigue in the newborn infant,” J. Pediatrics, vol. 95, pp. 793–797, 1979.
7
[10] S. Fucile, E. G. Gisel, and C. Lau, “Effect of an oral stimulation program on sucking skill maturation of preterm infants,” Develop. Med. Child Neurol., vol. 47, pp. 158–162, 2005. [11] J.-Y. Cheng, O. Amft, G. Bahle, and P. Lukowicz, “Designing sensitive wearable capacitive sensors for activity recognition,” IEEE Sensors J., vol. 13, pp. 3935–3947, 2013. [12] E. Sejdic', C. M. Steele, and T. Chau, “Classification of penetration–aspiration versus healthy swallows using dual-axis swallowing accelerometry signals in dysphagic subjects,” IEEE Trans. Biomed. Eng., vol. 60, pp. 1859–1866, 2013. [13] A. R. Fekr, M. Janidarmian, K. Radecka, and Z. Zilic, “A medical cloud-based platform for respiration rate measurement and hierarchical classification of breath disorders,” Sensors, vol. 14, pp. 11204–11224, 2014. [14] M. Krehel, M. Schmid, R. M. Rossi, L. F. Boesel, G. L. Bona, and L. J. Scherer, “An optical fibre-based sensor for respiratory monitoring,” Sensors, vol. 14, pp. 13088–13101, 2014. [15] Z. K. Moussavi, M. T. Leopando, H. Pasterkamp, and G. Rempel, “Computerised acoustical respiratory phase detection without airflow measurement,” Med. Biol. Eng. Comput., vol. 38, pp. 198–203, 2000. [16] L. J. Hadjileontiadis and I. T. Rekanos, “Detection of explosive lung and bowel sounds by means of fractal dimension,” IEEE Signal Process. Lett., vol. 10, pp. 311–314, 2003. [17] B. S. Lin, W. Chou, H. Y. Wang, Y. J. Huang, and J. S. Pan, “Development of novel non-contact electrodes for mobile electrocardiogram monitoring system,” IEEE J. Translational Eng. Health Med., vol. 1, pp. 1–8, 2013. [18] I. H. Gewolb, F. L. Vice, E. L. Schweitzer-Kenney, V. L. Taciak, and J. F. Bosma, “Developmental patterns of rhythmic suck and swallow in preterm infants,” Develop. Med. Child Neurol., vol. 43, pp. 22–27, 2001. [19] M. A. Qureshi, F. L. Vice, V. L. Taciak, J. F. Bosma, and I. H. Gewolb, “Changes in rhythmic suckle feeding patterns in term infants in the first month of life,” Develop. Med. Child Neurol., vol. 44, pp. 34–39, 2002. [20] E. S. Sazonov, O. Makeyev, S. Schuckers, P. Lopez-Meyer, E. L. Melanson, and M. R. Neuman, “Automatic detection of respiration rate from ambulatory single-lead ECG,” IEEE Trans. Inf. Technol. Biomed., vol. 13, pp. 890–896, 2009. [21] E. S. Sazonov, O. Makeyev, S. Schuckers, P. Lopez-Meyer, E. L. Melanson, and M. R. Neuman, “Automatic detection of swallowing events by acoustical means for applications of monitoring of ingestive behavior,” IEEE Trans. Biomed. Eng., vol. 57, pp. 626–633, 2009. [22] S. Damouras, E. Sejdic, C. M. Steele, and T. Chau, “An online swallow detection algorithm based on the quadratic variation of dual-axis accelerometry,” IEEE Trans. Signal Process., vol. 58, pp. 3352–3358, 2010. [23] C. Lau, “Oral feeding in the preterm infant,” NeoReviews, vol. 7, pp. 19–27, 2006. [24] L. Jain, E. Sivieri, S. Abbasi, and V. K. Bhutani, “Energetics and mechanics of nutritivesucking in the preterm and term neonate,” J. Pediatrics, vol. 111, pp. 894–898, 1987. [25] K. Mizuno and A. Ueda, “The maturation and coordination of sucking, swallowing, and respiration in preterm infants,” J. Pediatrics, vol. 142, pp. 36–40, 2003. [26] R. Paludetto, S. S. Robertson, and R. J. Martin, “Interaction between nonnutritive sucking and respiration in preterm infants,” Biol. Neonate, vol. 49, pp. 198–203, 1986. [27] P. H. Wolff, “The serial organization of sucking in the young infant,” Pediatrics, vol. 42, pp. 943–956, 1968.
Chang-Ting Chen is working toward the Ph.D. degree at the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu, Taiwan. His research interests are in the areas of biomedical circuits and systems, biomedical signal processing, and biosensors.
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 8
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
Yu-Lin Wang received the M.D. degree from Kaohsiung Medical University, Kaohsiung, Taiwan, in 1992. Since 1996, he has been a member of the visiting staff and Lecturer in the Rehabilitation Department of Kaohsiung Medical University Hospital and Chi Mei Medical Center, Tainan, Taiwan. His research activities involve sonograms image processing and electrophysiological signal processing.
Chen-An Wang is working toward the Master’s degree at the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu, Taiwan. Her research interests are in biomedical system design and signal processing.
Mei-Ju Ko received the M.S. degree from the Hearing and Speech Language Therapy Institute, National Kaohsiung Normal University, Kaohsiung, Taiwan, in 2012. Currently, she is the Speech-Language Therapist at Chi-Mei Medical Center, Tainan, Taiwan. Her specialty is in adult and children dysphagia.
Wai-Chi Fang (S’81–M’86–SM’93–F’03) received the B.Sc. degree from the Electronics Engineering Department at National Chiao Tung University, Hsinchu City, Taiwan, the M.Sc. degree from the State University of New York at Stony Brook, NY, USA, and the Ph.D. degree from the University of Southern California, Los Angeles, CA, USA, in 1978, 1982, and 1992, respectively. Currently, he is the TSMC Chair Professor at National Chiao Tung University. From 1985 to 2007, he was with NASA’s Jet Propulsion Laboratory (JPL), California Institute of Technology, Pasadena, CA, USA. His subjects of interest include VLSI bio-medical microsystems, neural networks and intelligent systems, multimedia signal processing, wireless communication, sensor networks, and space integrated avionic systems. He holds seven U.S. patents and 13 NASA new technologies. His inventions on advanced computing engines and data compression systems are used in space missions. Dr. Fang was the recipient of 1995 IEEE VLSI Transactions Best Paper Award and NASA Certificates of Recognition for these creative technical innovations. He won the NASA Space Act Award in 2002 and 2003. He is an elected Governor of the IEEE Circuits and Systems Society and an AdCom member of the IEEE Nanotechnology Council. He also serves as an officer of the IEEE Systems Council as the Chairman of Transnational and Liaison Committee. He is the current Chairman of the IEEE CASS Technical Committee on Nanoelectronics and Gigascale Systems. He serves on the Advisory Board of the IEEE Systems Journal and the Advisory Board of International Journal of Innovative Computing, Information and Control.
Bor-Shyh Lin (M’02–SM’15) received the B.S. degree from National Chiao Tung University (NCTU), Hsinchu City, Taiwan, in 1997, and the M.S. and Ph.D. degrees in electrical engineering from National Taiwan University, Taipei City, Taiwan, in 1999 and 2006, respectively. Currently, he is an Associate Professor at the Institute of Imaging and Biomedical Photonics at NCTU. His research interests are in the areas of biomedical circuits and systems, biomedical signal processing, and biosensors.
1
Wireless Monitoring System for Oral-Feeding Evaluation of Preterm Infants Chang-Ting Chen, Yu-Lin Wang, Chen-An Wang, Mei-Ju Ko, Wai-Chi Fang, Fellow, IEEE, and Bor-Shyh Lin, Senior Member, IEEE
Abstract—The oral feeding disorder is one of the important indicators for the high risk group of neurodevelopment delay. The procedure of oral feeding requires the coordination of sucking, swallowing, and breathing activities, and it is the most complex sensorimotor process for newborn infants. Premature infants often uneasily complete the procedure of oral feeding. However, the evaluation of the oral feeding disorders and severity are usually dependent on the subjective clinical experience of the physician. Monitoring the sucking-swallowing-breathing activities directly is difficult for preterm infants. In this study, a wireless monitoring system for oral-feeding evaluation of full term and preterm infants was proposed to objectively and quantitatively evaluate the coordination of suck-swallow-respiration function during oral feeding. Moreover, the ratios of the swallowing and breathing event numbers to the sucking event number were defined to evaluate the coordination of suck-swallow-respiration function during oral feeding. Finally, the system performance was validated and the coordination of suck-swallow-respiration function for full term and preterm infants during oral feeding was also investigated. Index Terms—Oral feeding disorder, premature infant, suck-swallow-respiration function, wireless monitoring system.
I. INTRODUCTION
P
REMATURE infants, that their gestation week is less than 37 week, usually have immature organs, and the immature organs always bring some risk factors. The oral feeding disorder is one of the common disease for premature infants. The procedure of oral feeding requires the precise coordination between sucking, swallowing and breathing activities. The premature infants may unsuccessfully complete the procedure
Manuscript received January 08, 2015; revised May 14, 2015; accepted May 23, 2015. The authors are greatly indebted to the Ministry of Science and Technology, Taiwan for the support of the research through contracts in 103-2221-E009-035-MY2. This paper was recommended by Associate Editor S. Sonkusale. C.-T. Chen and C.-A. Wang are with the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu 300, Taiwan. Y.-L. Wang is with the Department of Rehabilitation, Chi Mei Medical Center, Tainan 710, Taiwan, and also with the Department of Sports Management, Chia Nan University of Pharmacy and Science, Tainan City 717, Taiwan. M.-J. Ko is with the Department of Rehabilitation, Chi Mei Medical Center, Tainan 710, Taiwan. W.-C. Fang is with the Department of Electronics Engineering, National Chiao Tung University, Hsinchu 300, Taiwan. B.-S. Lin is with the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu 300, Taiwan, and also with the Department of Medical Research, Chi-Mei Medical Center, Tainan 710, Taiwan (e-mail: [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TBCAS.2015.2438031
of oral feeding, and this affects the absorption of nutrients and may result in bradycardia and apnea [1]. Many premature infants have to receive oral and non-oral sensorimotor interventions until these infants are able to independently complete nutritive sucking [2]. For preterm infants, oral motor intervention is a sensory stimulation of the lips, jaw, tongue, soft palate, pharynx, larynx and respiratory muscles pre-oral feeding, and assists to gain the rhythm of sucking-swallowing with monitoring stress signs (such as cyanosis, tachypnea, apnea or bradycardia) during oral feeding. The previous research has proved that the early oral motor intervention can shorten the transition time from gavage feeding to full oral feeding and improve oral feeding efficiency [3]. Non-oral sensorimotor intervention is a kind of tactile/ kinesthetic stimulations consisted of body stroking and passive movements of the limbs for three, 15-minute periods per day. The combined oral and non-oral sensorimotor intervention has an additive/synergistic effect for the independent complete nutritive sucking [4]. Preterm infants with heart or lung problems may need to breathe increased amounts of oxygen by the oxygen hood or nasal cannula to get normal levels of oxygen in their blood during oral feeding. Several previous studies further investigated the behavior of oral feeding for premature infants. Nutritive sucking is a variable process, and can be separated into the intermittent, continuous and pause phases during oral feeding from their sucking, swallowing, and breathing activities. During the first six months after birth, infants obtained their primary foods through nutritive sucking. By nutritive sucking, infants can be fed with lower energy requirement [5], [6]. However, the swallowing activity has to use laryngeal muscles, pharyngeal and oral simultaneously within a few milliseconds, and it is a difficult action for premature infants. Moreover, in order to prevent choking, the airways will be closed with the swallowing reflex motion [7]. Goldfield et al. indicated that the swallowing activity is not a random distribution during oral feeding procedure [8]. The event of swallowing activity happens before the next sucking event and between breathing out and breathing in. Infants cannot maintain ventilation, when the coordination of suck-swallow-respiration function collapses. Muller et al. investigated the weaning process of infants by using mechanical ventilators [9], and they indicated the infants’ diaphragmatic fatigue is highly related to their breathing activities. Fucile et al. investigated the suck-swallow-respiration function of the preterm infants during oral feeding, and they indicated that applying the oral sensory stimulation to infants can improve the ability of oral feeding [10]. Therefore, the parameters related to the coordination between sucking,
1932-4545 © 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 2
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
swallowing, and breathing activities are the key indicators for the oral feeding ability of premature infants. Up to now, the oral feeding ability is evaluated and diagnosed subjectively by medical staffs. Thus, developing the monitoring system to objectively and quantitatively evaluate the suck-swallow-respiration function of premature infants is important for the diagnosis of oral feeding disorders. Several methods have been proposed to monitor the swallowing or breathing activities. The accelerometer-based or sEMG-based approaches have been proposed to monitor the swallowing activity [11], [12]. Cheng et al. proposed a textile capacitive sensor to measure the capacitance variation of the human body and monitor swallowing activity [11]. Sejdic et al. used a dual-axis accelerometer to monitor the movement of the throat, and applied it to detect the state of dysphagic patients [12]. The approaches of using the acceleration [13], the respiratory belt [14], or microphone [15] have also been used to monitor the breathing activity in previous studies. However, premature infants are fragile and their muscle activities are unapparent. Therefore, the above methods are not suitable to monitor the swallowing and breathing activities of premature infants. Additionally, there is still lack of medical instruments for monitoring the sucking activity of infants directly during oral feeding due to that the electrical sensor is easily affected and destroyed by milk. In this study, a wireless monitoring system for oral-feeding evaluation of preterm infants was proposed to quantitatively evaluate the coordination of suck-swallow-respiration function during oral feeding. Here, a novel sensing device was designed and implemented to monitor the sucking pressure of infants during oral feeding. The raw surface electromyography (sEMG), measured from the chest of infants, and the swallowing sound were used to detect the events of breathing and swallowing activities respectively. Moreover, the ratios of the swallowing and breathing event numbers to the sucking event number were defined and used to evaluate the coordination of suck-swallow-respiration function during oral feeding. Finally, the system performance was validated and the coordination of suck-swallow-respiration function for premature infants and full-term infants during oral feeding was also investigated. II. SYSTEM ARCHITECTURE AND DESIGN The basic framework and photograph of the proposed wireless monitoring system for oral-feeding evaluation are shown in Fig. 1(a) and (b) respectively. The proposed system mainly contains a sensing device for monitoring sucking pressure, a wireless multi-parameter acquisition module, and a back-end system platform. Here, the nipple of the sensing device for sucking pressure was placed in the mouth of the infant to measure the sucking pressure under oral feeding, and convert the sucking pressure into electrical signal. The swallowing sound and the raw sEMG related to the breathing activity were acquired by a mini microphone and the Neonatal Electrodes (Kendall™ 1041PTS, Covidien, Ireland) respectively. Here, the mini microphone was placed on the neck (at the lower border of the cricoid cartilage), and a pair of surface electrodes were placed at the merging of the sixth and seventh intercostal space along the anterior axillary line with grounding electrode
Fig. 1. (a) Basic framework and (b) photograph of proposed wireless monitoring system.
at the xiphoid. Subsequently, these electrical signals of sucking pressure, swallowing sound and raw breathing sEMG were amplified, filtered, and digitized by the wireless multi-parameter acquisition module, and then were transmitted to the back-end system platform wirelessly via Bluetooth. Finally, the multi-parameter monitoring program built in the back-end system platform received these bio-signals and detected the events of sucking, swallowing, and breathing activities. The ratio of the swallowing event number to the sucking event number and the ratio of the breathing event number to the sucking event number were also calculated by the multi-parameter monitoring program to evaluate the ability of oral feeding for infants. A. Sensing Device for Monitoring Sucking Pressure Fig. 2 shows the basic structure and photograph of the sensing device for sucking pressure. It mainly consists of an electrical pressure sensor, a rubber tube and a polypropylene bottle. In this study, intra-nipple (teat) pressure was measured with the pressure sensor (SSC-SNBN400MD-AA3, Honey Well, U. S.), which was connected to plastic tubing positioned to lie within the feeding nipple teat. Its measuring range is within , and the range of operation temperature is from Celsius to 85 degrees Celsius. The pressure sensor contains two pressure inputs, and the pressure difference between the two pressure inputs will be convert into a voltage signal. However, it is difficult to measure the sucking pressure directly during oral feeding, because the milk easily
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. CHEN et al.: WIRELESS MONITORING SYSTEM FOR ORAL-FEEDING EVALUATION OF PRETERM INFANTS
Fig. 2. Basic structure and photograph of proposed sensing device for monitoring sucking pressure.
3
unit. Here, the front-end amplifier circuits consist of pre-amplifiers and band-pass filters. The operational amplifier (AD8609, Analog Devices. U.S) was used in the front-end amplifier. The gains of the pre-amplifiers and the frequency band of the band-pass filters were designed according to the characteristics of different bio-signals. The gains of the pre-amplifiers for swallowing sound and breathing sEMG were set to 900 times and 4000 times respectively. Here, the 2-order Butterworth band-pass filters were implemented. The frequency band of the band-pass filter for sucking pressure and breathing sEMG were set to 0.5 Hz – 500 Hz and 0.5 Hz – 500 Hz respectively, and the frequency band of the high-pass filter for swallowing sound was set to . After amplifying and filtering these bio-signals, a 12-bit analog-to-digital converter (ADC) built in the microprocessor was used to digitize these bio-signals. The sampling rates for sucking pressure, swallowing sound, and breathing sEMG were set to 1000 Hz simultaneously. In this study, the microprocessor (MSP430, Texas Instruments, U. S.) was used to control the analog-to-digital converter, package the data of these bio-signals, and transmit them to the wireless transmission circuit. The design of the wireless transmission circuit was based on a Bluetooth module, supporting the Bluetooth v2.0 + EDR specification, with a printed circuit board-based antenna. The power management unit provided the stable 3 V power supply and the function of charging the battery. Here, a commercial 250 mAh Li-ion battery was used in this module, and this module can be operated continuously for 9 hours. C. Back-End System Platform
Fig. 3. Block diagram and photograph of wireless multi-parameter acquisition module.
affects and destroys the electrical pressure sensor. In order to overcome the above issue, the electrical pressure sensor was placed at the outside of the polypropylene bottle, and the rubber tube was used to connect with the electrical pressure sensor and the polypropylene bottle. Here, one terminal of the rubber tube was connected with the input of the pressure sensor, and the other terminal was placed at the nipple inside of the polypropylene bottle. Based on the connected pipe, the intra-nipple (teat) pressure change during sucking motions can be transmitted to the input of the pressure sensor. By comparing with the general atmospheric pressure, the sucking pressure of the nipple during oral feeding was measured indirectly by the pressure sensor to avoid the influence of milk during oral feeding. B. Wireless Multi-Parameter Acquisition Module Fig. 3 illustrates the block diagram and photograph of the wireless multi-parameter acquisition module. The size of the proposed module is about 8.5 5 1.8 cm3. It mainly contains several parts: front-end amplifier circuits, a microprocessor unit, a wireless transmission circuit, and a power management
The design of the back-end system platform in this system was based on a commercial laptop, and Windows 8.1 was used as the operation system of the system platform. A multi-parameter monitoring program built in the back-end system platform was developed by Microsoft C#, and it can provide the basic functions of receiving, displaying, and storing raw data, detecting the event number of sucking, swallowing, and breathing activities, and calculating the ratios of the event numbers of swallowing and breathing activity to that of sucking activity. III. SYSTEM SOFTWARE DESIGN The operation procedure and the screenshot of the multi-parameter monitoring program are shown in Fig. 4(a) and (b) respectively. The multi-parameter monitoring program mainly consists of three parts: GUI, BUFFER, and THREAD. Here, BUFFER is a container used to store data received from the wireless multi-parameter acquisition module. The part of GUI provided the ability to precisely control the location and display of the GUI elements, such as the form, panel, et al. The part of THREAD denotes the execution thread, and it contains the threads of BT API, REC, and ANALYSIS. Here, the thread of BT API and REC were designed to connect with the wireless multi-parameter acquisition module via Bluetooth, and receive the raw data and store them in BUFFER respectively. The design of the ANALYSIS thread was based on the multi-parameter analysis algorithm, which was designed to detect the event number of sucking, swallowing, and breathing activities, and the ratios of the event numbers of swallowing and breathing
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 4
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
Fig. 4. (a) Operation procedure and (b) screenshot of multi-parameter monitoring program.
activity to that of sucking activity to estimate the coordination of suck-swallow-respiration function. The procedure of the multi-parameter analysis algorithm is illustrated in Fig. 5. First, the raw signals of swallowing sound and breathing sEMG was pre-processed by high-pass filters to remove the low frequency noise and the interference of electrocardiogram (ECG). Next, the technique of fractal dimension (FD) [16] was applied to estimate the variation of signal complexity for the swallowing sound and breathing sEMG. The variation of the signal complexity was used as the event index of the swallowing and breathing activities. The fractal dimension value can be calculated by
(1) where denotes the total length of the curve, is the largest distance between the first point and any other point on the curve. Here, is defined as the number of steps in the curve, and can be calculated by (2) is the average distance between each consecutive where points. Here, a two second moving window with the overlap of one second was used to calculate the fractal dimension value. When the activity event occurred, the signal complexity of the biosignal increased to result in the increase of the fractal dimension value. Therefore, the events of the swallowing
Fig. 5. Flowchart of multi-parameter analysis algorithm.
and breathing activities reflected on the change of the fractal dimension value indirectly. Next, the positive peaks of the sucking pressure, and the fractal dimension values of the swallowing sound and breathing sEMG were detected by using the first derivative (FDI) approach [17], and these positive
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. CHEN et al.: WIRELESS MONITORING SYSTEM FOR ORAL-FEEDING EVALUATION OF PRETERM INFANTS
5
TABLE I PERFORMANCE OF DETECTING SUCKING EVENTS BY PROPOSED ALGORITHM
PERFORMANCE
TABLE II DETECTING SWALLOWING EVENTS PROPOSED ALGORITHM
OF
BY
TABLE III PERFORMANCE OF DETECTING BREATHING EVENTS BY PROPOSED ALGORITHM
Fig. 6. Number of sucking, swallowing, and breathing events corresponding to different infant ages.
The sensitivity and PPV for detecting the event of the swallowing activity were 92.89% and 91.31%. The sensitivity and PPV for detecting the event of the breathing activity were 96.51% and 87.18%. peaks were viewed as the events of sucking, swallowing, and breathing activities. Finally, the event numbers of sucking, swallowing, and breathing activities per one minute were used to calculate the ratio and to estimate the coordination of suck-swallow-respiration function during oral feeding. Here, the parameters and denote the ratio of the swallowing event number to the sucking event number and the ratio of the breathing event number to the sucking event number respectively. IV. RESULTS A. Performance of Multi-Parameter Analysis Algorithm In this section, the performance of the proposed multi-parameter analysis algorithm was investigated. Here, several types of binary classification test parameters for validating the algorithm performance were first defined, and listed as follows: true positive (TP) means the activity event is correctly detected as an event, false positive (FP) means that no activity event is wrongly detected as an event, false negative (FN) means that the activity event is wrongly detected as nothing, and true negative (TN) means that no activity event is correctly detected as nothing. Tables I–III show the performance of detecting the events of the sucking, swallowing, and breathing activities by using the proposed algorithm respectively. For detecting the event of the sucking activity, the sensitivity and positive predictive value (PPV) were 96.7% and 94.25%. Here, the positive predictive value denotes the proportions of positive results in statistics that are true positive results and sensitivity denotes the proportion of actual positives which are correctly identified as positives, and they can be given by (3) (4)
B. Coordination Evaluation of Suck-Swallow-Respiration Function In this section, the coordination of suck-swallow-respiration function for preterm infants and full-term infants were evaluated. There were 22 for full-term ( ) Asian infants, 26 for 34 weeks, 9 for 35 weeks and 5 for 36 weeks of postmenstrual age premature Asian infants that attended this experiment. This experiment has been approved by the Institutional Review Board (IRB), Chi-Mei medical center, Taiwan, and all participants provided an informed consent. Fig. 6 shows the event number of sucking, swallowing, and breathing activities corresponding to different infant ages. The event numbers of the sucking and swallowing activities increased with the increase of the infant age. However, the event number of the breathing activity decreased with the increase of the infant age. For the full-term infants, the ratio of the sucking, swallowing, and breathing activities was about 1:1:1. Fig. 7 shows the ratios and for the infants with different ages. The analysis of variance (ANOVA) method was used for the analysis, and the difference significance was defined as in this study. Both of the ratio and for preterm infants were significantly different from that of full-term infants. The ratio was close to 1, and its standard deviation decreased with the increase of the infant age. The ratio of decreased obviously and became more stable with the increase of the infant age. For the full-term infant, both of the ratio and became close to 1, and were relatively stable. V. DISCUSSIONS Several monitoring systems have been developed in previous studies to detect swallowing and breathing activities. Table IV shows the system comparison between the proposed system and other monitoring systems. Different from the neck of the adult, the neck of the premature infant is very short, and the muscle movement of the infant is unapparent when
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 6
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
TABLE IV SYSTEM COMPARISON BETWEEN PROPOSED SYSTEM AND OTHER MONITORING SYSTEMS
Fig. 7. Ratios of RSS and RBS corresponding to different infant ages. Here, the sign * denotes the significant difference.
swallowing. Attaching the surface electrodes to the infant’s neck is difficult and inconvenient, and the swallowing sEMG is also easily affected by other biological potentials. By replacing the accelerometer-based or sEMG-based approaches used in the previous studies, the proposed system used the microphone to successfully monitor the swallowing activity. Moreover, by replacing the approaches of using the respiratory belt and the acceleration, the proposed system used sEMG to successfully monitor the breathing activity, and improve the issue of the chest movement of infants is very small when breathing. The sensing device for monitoring sucking pressure was also designed to successfully monitor the intra-nipple (teat) pressure. By the special mechanical design of the sensing device, it successfully avoided the influence of milk on the electrical pressure sensor [18], [19]. Several algorithms have also been developed to detect the breathing and swallowing activities in previous studies. In 2009, Sazonov et al. proposed the ECG-derived respiration technique which was based on the technique of wavelet decomposition to extract the information of breathing activity
from one lead ECG signal [20], and its mean error rate was about . In 2009, Sazonov et al. used the approach of support vector machine with mel-scale Fourier spectrum and wavelet packet decomposition techniques to detect swallowing activity, and its accuracy of detecting swallowing activity was about 84.7% [21]. In 2010, Damouras et al. proposed the swallow detection algorithm based on the quadratic variation technique to detect swallowing activity from the dual-axis accelerometry signal, and its accuracy was more than 90% [22]. By using the technique of fractal dimension, the events of swallowing and breathing activities were exactly reflected on the change of their fractal dimension values. Most of the sucking, swallowing, and breathing events were effectively detected by the proposed algorithm. From the experimental results, the state of false positive usually occurred when a longer and continuous activity event was detected as a number of separate activity events. The state of false negative usually occurred when a short or unapparent activity event happened. In this study, the ratios of and were also defined to evaluate the coordination of suck-swallow-respiration function for infants. Both of the ratio and became smaller and more stable with the increase of the infant age. From the concept of cross-systems interactions, in order to ensure the infant feeding safely, central pattern generators (CPGs) integrate and coordinate the motor neurons of swallowing, sucking, and breathing [18]. With the gradual maturation of the infant, the sucking and swallowing abilities of the infant become more proficient, and this also results in that the sucking and swallowing events become more rapid and coordinated. In general, the ratio of the sucking events to the swallowing event for a healthy full-term infant is about 1:1 or 2:1 [23]. Therefore, the ratio of was close to 1, and became more stable when the infant age increased. The rate of respiratory activity is also an important indicator to evaluate the ability of oral feeding for infants. During the continuous sucking phase of oral feeding, the rate of respiratory activity for the preterm infants is about 26–31 events per minute, and that for full-term infants is about 30–35
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. CHEN et al.: WIRELESS MONITORING SYSTEM FOR ORAL-FEEDING EVALUATION OF PRETERM INFANTS
events per minute [24], [25]. The preterm infant needs more prolonged breath in the alternating periods of sucking, swallowing, and breathing activities. Because the prolonged breath pauses in the alternating periods or blocking the nipple with the tongue, the preterm infants have poor and irregular coordination of sucking, swallowing, and breathing activities during oral feeding. When the preterm infants matures, the ratio of sucking events to breathing events will also trend to 1:1 [26], [27]. Therefore, the ratio of also became close to 1 and more stable when the infant age increased.
VI. CONCLUSIONS In this study, a wireless monitoring system was proposed to continuously and non-invasively monitor the sucking-swallowing-breathing activities during oral feeding and evaluate the ability of oral feeding for premature infants. The sensing device for sucking pressure was also successfully applied in monitoring the sucking pressure during oral feeding to avoid the influence of milk. From the experimental results, the events of sucking-swallowing-breathing activities were detected effectively by the proposed multi-parameter analysis algorithm. Moreover, the ratios of the swallowing and breathing event number to the sucking event number were also defined to evaluate the coordination of suck-swallow-respiration function for infants. The ratio of for premature infants was significantly larger than that of full-term infants. Both of the ratio and became stable with the increase of the infant age. For full-term infants, the ratio of sucking, swallowing, and breathing was close to 1:1:1.
REFERENCES [1] C. Slocum, M. Arko, J. Di Fiore, R. J. Martin, and A. M. Hibbs, “Apnea, bradycardia and desaturation in preterm infants before and after feeding,” J. Perinatol., vol. 29, pp. 209–212, 2009. [2] B. Medoff-Cooper and W. Ray, “Neonatal sucking behaviors,” J. Nursing Scholar., vol. 27, pp. 195–200, 1995. [3] J. Arvedson, H. Clark, C. Lazarus, T. Schooling, and T. Frymark, “Evidence-based systematic review: Effects of oral motor interventions on feeding and swallowing in preterm infants,” Amer. J. Speech-Language Pathol., vol. 19, no. 4, pp. 321–340, 2010. [4] S. Fucile, E. G. Gisel, D. H. McFarland, and C. Lau, “Oral and non-oral sensorimotor interventions enhance feeding performance in preterm infants,” Develop. Med. Child Neurol., vol. 53, pp. 829–835, 2011. [5] J. Koenig, A. Davies, and B. Thach, “Coordination of breathing, sucking, and swallowing during bottle feedings in human infants,” J. Appl. Phys., vol. 69, pp. 1623–1629, 1990. [6] B. Medoff-Cooper, W. Bilker, and J. M. Kaplan, “Sucking patterns and behavioral state in 1-and 2-day-old full-term infants,” J. Obstetric, Gynecol., Neonatal Nursing, vol. 39, pp. 519–524, 2010. [7] J. F. Bosma, “Postnatal ontogeny of performances of the pharynx, larynx, and mouth,” Amer. Rev. Respiratory Disease, vol. 131, pp. 10–15, 1985. [8] E. C. Goldfield, C. Buonomo, K. Fletcher, J. Perez, S. Margetts, A. Hansen, V. Smith, S. Ringer, M. J. Richardson, and P. H. Wolff, “Premature infant swallowing: Patterns of tongue-soft palate coordination based upon videofluoroscopy,” Infant Behav. Develop., vol. 33, pp. 209–218, 2010. [9] N. Muller, G. Volgyesi, M. H. Bryan, and A. C. Bryan, “The consequences of diaphragmatic muscle fatigue in the newborn infant,” J. Pediatrics, vol. 95, pp. 793–797, 1979.
7
[10] S. Fucile, E. G. Gisel, and C. Lau, “Effect of an oral stimulation program on sucking skill maturation of preterm infants,” Develop. Med. Child Neurol., vol. 47, pp. 158–162, 2005. [11] J.-Y. Cheng, O. Amft, G. Bahle, and P. Lukowicz, “Designing sensitive wearable capacitive sensors for activity recognition,” IEEE Sensors J., vol. 13, pp. 3935–3947, 2013. [12] E. Sejdic', C. M. Steele, and T. Chau, “Classification of penetration–aspiration versus healthy swallows using dual-axis swallowing accelerometry signals in dysphagic subjects,” IEEE Trans. Biomed. Eng., vol. 60, pp. 1859–1866, 2013. [13] A. R. Fekr, M. Janidarmian, K. Radecka, and Z. Zilic, “A medical cloud-based platform for respiration rate measurement and hierarchical classification of breath disorders,” Sensors, vol. 14, pp. 11204–11224, 2014. [14] M. Krehel, M. Schmid, R. M. Rossi, L. F. Boesel, G. L. Bona, and L. J. Scherer, “An optical fibre-based sensor for respiratory monitoring,” Sensors, vol. 14, pp. 13088–13101, 2014. [15] Z. K. Moussavi, M. T. Leopando, H. Pasterkamp, and G. Rempel, “Computerised acoustical respiratory phase detection without airflow measurement,” Med. Biol. Eng. Comput., vol. 38, pp. 198–203, 2000. [16] L. J. Hadjileontiadis and I. T. Rekanos, “Detection of explosive lung and bowel sounds by means of fractal dimension,” IEEE Signal Process. Lett., vol. 10, pp. 311–314, 2003. [17] B. S. Lin, W. Chou, H. Y. Wang, Y. J. Huang, and J. S. Pan, “Development of novel non-contact electrodes for mobile electrocardiogram monitoring system,” IEEE J. Translational Eng. Health Med., vol. 1, pp. 1–8, 2013. [18] I. H. Gewolb, F. L. Vice, E. L. Schweitzer-Kenney, V. L. Taciak, and J. F. Bosma, “Developmental patterns of rhythmic suck and swallow in preterm infants,” Develop. Med. Child Neurol., vol. 43, pp. 22–27, 2001. [19] M. A. Qureshi, F. L. Vice, V. L. Taciak, J. F. Bosma, and I. H. Gewolb, “Changes in rhythmic suckle feeding patterns in term infants in the first month of life,” Develop. Med. Child Neurol., vol. 44, pp. 34–39, 2002. [20] E. S. Sazonov, O. Makeyev, S. Schuckers, P. Lopez-Meyer, E. L. Melanson, and M. R. Neuman, “Automatic detection of respiration rate from ambulatory single-lead ECG,” IEEE Trans. Inf. Technol. Biomed., vol. 13, pp. 890–896, 2009. [21] E. S. Sazonov, O. Makeyev, S. Schuckers, P. Lopez-Meyer, E. L. Melanson, and M. R. Neuman, “Automatic detection of swallowing events by acoustical means for applications of monitoring of ingestive behavior,” IEEE Trans. Biomed. Eng., vol. 57, pp. 626–633, 2009. [22] S. Damouras, E. Sejdic, C. M. Steele, and T. Chau, “An online swallow detection algorithm based on the quadratic variation of dual-axis accelerometry,” IEEE Trans. Signal Process., vol. 58, pp. 3352–3358, 2010. [23] C. Lau, “Oral feeding in the preterm infant,” NeoReviews, vol. 7, pp. 19–27, 2006. [24] L. Jain, E. Sivieri, S. Abbasi, and V. K. Bhutani, “Energetics and mechanics of nutritivesucking in the preterm and term neonate,” J. Pediatrics, vol. 111, pp. 894–898, 1987. [25] K. Mizuno and A. Ueda, “The maturation and coordination of sucking, swallowing, and respiration in preterm infants,” J. Pediatrics, vol. 142, pp. 36–40, 2003. [26] R. Paludetto, S. S. Robertson, and R. J. Martin, “Interaction between nonnutritive sucking and respiration in preterm infants,” Biol. Neonate, vol. 49, pp. 198–203, 1986. [27] P. H. Wolff, “The serial organization of sucking in the young infant,” Pediatrics, vol. 42, pp. 943–956, 1968.
Chang-Ting Chen is working toward the Ph.D. degree at the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu, Taiwan. His research interests are in the areas of biomedical circuits and systems, biomedical signal processing, and biosensors.
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 8
IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS
Yu-Lin Wang received the M.D. degree from Kaohsiung Medical University, Kaohsiung, Taiwan, in 1992. Since 1996, he has been a member of the visiting staff and Lecturer in the Rehabilitation Department of Kaohsiung Medical University Hospital and Chi Mei Medical Center, Tainan, Taiwan. His research activities involve sonograms image processing and electrophysiological signal processing.
Chen-An Wang is working toward the Master’s degree at the Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Hsinchu, Taiwan. Her research interests are in biomedical system design and signal processing.
Mei-Ju Ko received the M.S. degree from the Hearing and Speech Language Therapy Institute, National Kaohsiung Normal University, Kaohsiung, Taiwan, in 2012. Currently, she is the Speech-Language Therapist at Chi-Mei Medical Center, Tainan, Taiwan. Her specialty is in adult and children dysphagia.
Wai-Chi Fang (S’81–M’86–SM’93–F’03) received the B.Sc. degree from the Electronics Engineering Department at National Chiao Tung University, Hsinchu City, Taiwan, the M.Sc. degree from the State University of New York at Stony Brook, NY, USA, and the Ph.D. degree from the University of Southern California, Los Angeles, CA, USA, in 1978, 1982, and 1992, respectively. Currently, he is the TSMC Chair Professor at National Chiao Tung University. From 1985 to 2007, he was with NASA’s Jet Propulsion Laboratory (JPL), California Institute of Technology, Pasadena, CA, USA. His subjects of interest include VLSI bio-medical microsystems, neural networks and intelligent systems, multimedia signal processing, wireless communication, sensor networks, and space integrated avionic systems. He holds seven U.S. patents and 13 NASA new technologies. His inventions on advanced computing engines and data compression systems are used in space missions. Dr. Fang was the recipient of 1995 IEEE VLSI Transactions Best Paper Award and NASA Certificates of Recognition for these creative technical innovations. He won the NASA Space Act Award in 2002 and 2003. He is an elected Governor of the IEEE Circuits and Systems Society and an AdCom member of the IEEE Nanotechnology Council. He also serves as an officer of the IEEE Systems Council as the Chairman of Transnational and Liaison Committee. He is the current Chairman of the IEEE CASS Technical Committee on Nanoelectronics and Gigascale Systems. He serves on the Advisory Board of the IEEE Systems Journal and the Advisory Board of International Journal of Innovative Computing, Information and Control.
Bor-Shyh Lin (M’02–SM’15) received the B.S. degree from National Chiao Tung University (NCTU), Hsinchu City, Taiwan, in 1997, and the M.S. and Ph.D. degrees in electrical engineering from National Taiwan University, Taipei City, Taiwan, in 1999 and 2006, respectively. Currently, he is an Associate Professor at the Institute of Imaging and Biomedical Photonics at NCTU. His research interests are in the areas of biomedical circuits and systems, biomedical signal processing, and biosensors.
