Yiwey.net

Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 Journal of Zhejiang University-SCIENCE B (Biomedicine & Biotechnology) ISSN 1673-1581 (Print); ISSN 1862-1783 (Online) www.zju.edu.cn/jzus; www.springerlink.com E-mail: [email protected] Gastric motility functional study based on electrical bioimpedance
measurements and simultaneous electrogastrography*
Zhang-yong LI1, Chao-shi REN†‡1, Shu ZHAO2, Hong SHA2, Juan DENG2 (1College of Bioinformation, Chongqing University of Posts and Telecommunications, Chongqing 400065, China) (2Institute of Biomedical Engineering, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin 300192, China) †E-mail: [email protected] Received Dec. 10, 2010; Revision accepted Apr. 18, 2011; Crosschecked Nov. 7, 2011 Abstract: For some time now, the research on gastric motility and function has fallen behind in the amount of re-
search on gastric endocrine, exocrine secretion, and gastric morphology. In this paper, a noninvasive method to study
gastric motility was developed, taking bioimpedance measurements over the gastric area simultaneously with the
electrogastrography (EGG). This is based on the concept of observing and analyzing simultaneously the intrinsic
electrical gastric activity (basic electric rhythm) and the mechanical gastric activity. Additionally, preliminary clinical
studies of healthy subjects and subjects with functional dyspepsia (FD) and gastritis were carried out. The impedance
gastric motility (IGM) measurements of the healthy and FD subjects were compared, along with the studies of the FD
subjects before treatment and after one week and three weeks of treatment. We also compared IGM measurements of
healthy subjects and subjects with erosive gastritis, along with the studies of the subjects with erosive gastritis before
treatment and after one week of treatment. Results show that FD subjects have poor gastric motility (P<0.01). After a
week of treatment, the gastric motility of FD subjects was not yet improved although the EGG had returned to normal
by this time. By three weeks of treatment, the regular IGM rhythm returned in FD subjects. There was a significant
difference of IGM parameters between the gastritis and healthy subjects (P<0.05). The EGG rhythm of the gastritis
subjects returned to normal at one week post-treatment, while IGM parameters showed a trend to improvement
(P>0.05), These results suggest the possibility of clinic application of the proposed method.
Key words: Gastric motility, Electrical bioimpedance, Electrogastrography, Functional dyspepsia, Gastritis
doi:10.1631/jzus.B1000436 Document code: A CLC number: R31

1 Introduction
and function has fallen behind in the amount of re- search on gastric endocrine, exocrine secretion, and Gastric motility is one of the most critical phy- gastric morphology. One of the important reasons is siological functions of the human body. Without co- an absence of convenient and effective measurement ordinated gastric motility, digestion and absorption of methods (Zhou and Ke, 2005). dietary nutrients cannot take place. Impairment in Electrical bioimpedance measurements are de- gastric motility results in delayed emptying of the pendent on the electrical properties of tissues and stomach and symptoms, such as nausea, vomiting, organs, and include morphological and functional and abdominal pain or discomfort (Chen et al., 2000). information. The method has the outstanding advan- For some time now, the research on gastric motility tages of being noninvasive, convenient, and providing considerable functional information. It can be con- sidered a powerful tool in clinical diagnosis and ‡ Corresponding author medical research (Gajre et al., 2006). There are many * Project (Nos. 60471041 and 60901045) supported by the National applications using bioimpedance signals for different Natural Science Foundation of China Zhejiang University and Springer-Verlag Berlin Heidelberg 2011 pathological conditions, but their use in gastric Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 motility assessment needs to be explored in detail 2 Methods
(Hadi et al., 2002).
2.1 Composite course from electrical gastric ac-
Sutton and Thompson (1985) reported their re- tivity to mechanical gastric activity
search on extracting gastric movement signal by the electrical impedance method, and a curve reflecting Gastric contraction is a mechanical behavior of gastric emptying was obtained. From the curve, gas- the electrical activity occurring on the cell membrane tric peristaltic information with a rhythm 2–4 cycles surface of the smooth muscle. It begins from electric per min (cpm), which is in accord with gastric con- activity of the smooth muscle, followed by evoked traction, was extracted. Familoni et al. (1987) pre- contraction of the gastric corpus and antrum, and then sented a technique to monitor gastric electrical activ- transmits to the distal pylorus. It is a composite course ity (GEA) and mechanical activity as an aid in as- from electrical activity to mechanical contraction, sessing gastric motor function. Kothapalli (1992) then to gastric peristalsis and transmission. Gastric established a three-dimensional (3D) abdomen model contraction complies with the rhythm of electric ac- to study the origin of changes in the epigastric signal, tivity, and is affected by amplitude, time limitation, and analyzed the relationship between gastric im- transmission direction, and distance of the transmis- pedance signal and food capacity, resistivity of the sion contraction (Zhou and Ke, 2005). Gastric motil-test meal, and gastric contraction when the excitation ity is a complex composite course from electrical gas- electrode and measurement electrode were located at tric activity to mechanical gastric activity, and it is very different positions. important to measure and evaluate gastric motility Early research with the impedance method to according to the composite course (Ren et al., 2010). measure digestion was mainly concentrated on gastric There are two kinds of gastric myoelectrical ac- emptying measurement (Chaw et al., 2001; Giou- tivity to be observed, the slow wave and the spike vanoudi et al., 2003; Huerta-Franco et al., 2009). potential. Gastric antrum contraction occurs only There is little research on extraction of gastric motil- when the slow wave occurs with the spike potential. ity information (Garay et al., 2006; Soulsby et al., The spike potential appears during the slow wave 2006; Giouvanoudi and Spyrou, 2008). One of the phase, and the rhythm of gastric contraction may be primary reasons is that the rhythm of gastric motility determined by the slow wave (Ma et al., 2006). The is much lower, about 3 cpm. It is more difficult to EGG recorded from the body surface reflects the extract the gastric motility signal and eliminate res- myoelectrical activity of different areas of the stom- piration interference. Chen and Wan (1992b) reported ach, but corresponds to gastric slow wave accurately, their work on obtaining the electric impedance signal and therefore can be used to investigate the rhythm of to reflect gastric contraction, and measurement de- gastric contraction. vices were developed (Chen and Wan, 1992a). Stomach volume augments gradually when food However, the impedance signals obtained by the de- is ingested. In the gastric active period, such as con- vices are all similar sine waveforms for both healthy traction and peristalsis after a meal, the content of the and diseased subjects, because of incorrect filter stomach changes greatly, as does the impedance of processing. It has been difficult to differentiate nor- the stomach. Via the impedance measurement of the mal or abnormal conditions from the signals. We have stomach during digestion, the information reflecting proposed a noninvasive electrical impedance method the stomach volume (gastric emptying) and gastric for gastric motility measurement and evaluation pre- motility (contraction and peristalsis) can be extracted. viously (Li et al., 2007). Multi-resolution analysis of 2.2 Experimental procedures of IGM and EGG
the wavelet is adopted to separate impedance gastric motility (IGM) signal from the mixed impedance We have developed a study method of gastric signal obtained on the body surface (Li and Ren, motility based on electrical bioimpedance measure-2008). In this paper, we take bioimpedance meas- ments and simultaneous EGG (Li et al., 2007; 2008; urements with simultaneous electrogastrography Deng et al., 2008). The block diagram of the meas-(EGG) to study gastric motility. urement system is shown in Fig. 1. Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 rhythm belong to the ultra-low frequency signal and Sine signal generator the amplitude of the breath signal is usually much higher than that of IGM. It is a challenge to extract IGM information effectively from the mixed signal. A low-pass filter may eliminate the influences of Demodulation/filter high-frequency noise and heart activity interference; however, it is difficult to eliminate the influence of Computer/software Data acquisition respiration and separate IGM signal from the mixed signal. Thus, one uses a narrow band-pass filter and Fig. 1 Diagram of gastric motility measurement system
the high-order active low-pass filter. In the gastric I1, I 2: excitation electrodes; V1, V2: measurement electrodes motility measurement system, the wavelet transform is introduced and then IGM signal is separated suc- The measurement system consists of a sine sig- cessfully from the impedance signals of breath and nal generator, current source, electrodes (I1, I2, V1, blood flow among the mixed signal (Li et al., 2007). V2), preamplifier, demodulation/filter circuit, data The rhythm of the IGM signal is classified. The acquisition system, and a computer. A measurement rhythm of 2–4 cpm is the normal rhythm, while that current of 50 kHz, 2 mA provided by the current below 2 cpm is bradygastria, and that above 4 cpm is source goes into the abdomen zone of the measured tachygastria. Based on this classification, we carried subject via excitation electrodes I1 and I2 (Fig. 2). The out analyses of frequency spectra, energy spectra, impedance signal picked up from measurement elec- dynamic spectra, running spectra, frequency instabil- trodes V1 and V2 is fed into preamplifier, then the ity coefficient (FIC), power instability coefficient demodulation/filter, and goes into data acquisition (PIC), percentage of normal frequency (PNF), and system where analog-to-digital (A/D) conversion and percentage of normal power (PNP), for both IGM and digitization processing are carried out. The digital EGG. data then is sent to a computer where the proprietary The definitions of FIC and PIC are as below (Li software is in charge of the IGM and EGG informa- tion extraction, analyzing, and parameter calculations. Gastric emptying measurement and IGM and EGG spectral analyses are also executed by the proprietary software in the computer. SD = (DF(k) − DF ) S , where DFAv is the average of the dominant frequency within 2–4 cpm among the signal segments analyzed, SDF is the standard deviation of the dominant fre- quency, DF(k) is the dominant frequency for the kth signal segment analyzed, DFAvg is the average of the dominant frequency for all signal segments analyzed, and S is the number of all the segments. Fig. 2 Position of the excitation and measurement
elec trodes
The impedance signal acquired from the abdo- SD = (DP(k) − DP ) S , men surface is a mixed signal. It contains not only IGM, but also the components of impedance blood flow, breath, and some other disturbances. The nor- where DPAv is average of the dominant power within mal rhythm of IGM is about 3 cpm and the breath 2–4 cpm among the signal segments analyzed, SDP is signal is about 12 cpm. Both IGM and the breath the standard deviation of the dominant power, DP(k) Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 is the dominant power for the kth signal segment 3.2 Gastric motility measurement of FD patients
analyzed, DPAvg is the average of the dominant power
for all signal segments analyzed, and S is the number Gastric motility measurements of 30 healthy of all the segments. volunteers (control group) and 28 FD subjects (study group) were carried out. The volunteers were univer- 2.3 Body measurement and statistical analysis
sity teachers with ages of (45.2±12.3) years. Subjects The measurement method of gastric motility with FD (aged (40.9±9.7) years) came from the First used in the study is a noninvasive method. The study Affiliated Hospital of Chongqing University of was approved by the ethical committee, and all sub- Medical Sciences, and were diagnosed according to jects in the study signed a consent form. the Rome III classification for FD. Results of the IGM In preparation for the measurement, the subject measurement are shown in Table 1. It can be seen was calm and seated on a chair. The signals of IGM from Table 1 that the PNF in the study group (FD and EGG were recorded continually for 30–40 min subjects) is obviously lower than that in the control after a test meal. The meal consists of milk (200 ml) group (P<0.01), and the FIC is higher than that in the and bread (100 g), 1300 and 2850 kJ, respectively. control group (P<0.01). The PNP and PIC between The statistical software SPSS 13.0 was used to the two groups are also different (P<0.01). This in- analyze the data. The data are expressed as mean±SD. dicates that the FD subjects have a poor rhythm of Variance analysis between the study and control gastric motility. The results are in accordance to gas-groups was undertaken, and the significant difference tric physiology and pathology principles for FD. was accepted when P<0.05. 3 Results
In order to validate the feasibility and show the potential applications of the proposed method in this paper, some preliminary studies of gastric motility measurements in healthy volunteers, subjects with functional dyspepsia (FD), and subjects with gastritis have been carried out (Li and Ren, 2009; Liu et al., 3.1 Gastric motility measurement of the healthy
subjects
Fig. 3 IGM measurement and spectral analysis result for
The raw IGM signal and the spectral analyses of a healthy volunteer
the IGM and EGG signals for a healthy volunteer are (a) Original mixed impedance signal which comes from the illustrated in Fig. 3. The dynamic spectrum in Fig. 3d abdomen surface; (b) IGM signal extracted from the mixed signal; (c) Synchronous EGG; (d) The dynamic spectra of IGM shows that the dominant frequency of IGM signals (the interval of each spectrum line is 1 min); (e, f) Power spectra focuses on 2.8 cpm. before and after the test meal, respectively Table 1 IGM measurement results of healthy and FD subjects
Control 30 68.47±26.43 60.13±3.34 1.36±0.08 0.18±0.16 Study 28 28.32±16.92 50.79±9.90 2.08±0.55 0.23±0.05 Data are expressed as mean±SD. PNF: percentage of normal frequency; PNP: percentage of normal power; FIC: frequency instability coef- ficient; PIC: power instability coefficient

Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 Fig. 4 shows the EGG dynamic spectra of FD and compared the IGM and EGG parameters for the subjects before treatment and in a week treatment gastritis subjects before and after one week of treat-(domperidone tablets, 10 mg, three times per day for ment (sodium rabeprazole tablets, 10 mg, one time oral administration, half an hour before meals), and per day for oral administration in the morning). Thirty the interval of each spectrum line is 1 min. In Fig. 4, healthy volunteers (control group) were university EGG of FD patient is weak and the rhythm is disor- teachers with ages of (45.2±12.3) years. Thirty sub- dered before the treatment (Fig. 4a); then in a week jects with erosive gastritis (study group; aged treatment, the EGG enhances and the rhythm gets (50.5±13.0) years) came from the First Affiliated back to 2–4 cpm (Fig. 4b). This suggests that the Hospital of Chongqing University of Medical Sci-gastric electric activity of FD patients tended towards ences and were diagnosed by gastroscope examina- normal after a week treatment. tion. The statistic results of the studies are shown in Fig. 5 shows the IGM signal spectra of FD sub- Tables 2, 3, and 4. jects before treatment, after one week of treatment, Table 2 shows that there is a significant differ- and after three weeks of treatment. Compared with ence in the IGM parameters between the control and the EGG dynamic spectra in Fig. 4, the IGM in Fig. 5 study groups. PNF and PNP of the study group are shows little change after one week of treatment, and significantly lower than those of the control group the rhythm also is disordered; until after three weeks (P<0.05). FIC and PIC of the study group are evi- of treatment, the IGM rhythm appears back to 2–4 cpm. dently higher than those of the control group (P<0.05). These results suggest that the gastric motility function 3.3 Gastric motility evaluation of gastritis patients
of the gastritis subjects was reduced and the stomach We conducted IGM parameter measurement peristalsis was disordered. of healthy volunteers and erosive gastritis subjects, Table 3 indicates that the EGG power of normal rhythm (2–4 cpm) for the gastritis subjects was raised and the power of the abnormal rhythm (0–2 cpm and >4 cpm) declined after one week of treatment, al- though this was not statistically significant (P>0.05). It is important that FIC and PIC of the gastritis sub- jects were reduced (P<0.05). This suggests that the EGG of the gastritis subjects tended to be regular and stable after one week of treatment, after which the EGG rhythm improved. The IGM parameters in Table 4 show that the of IGM signals for the gastritis subjects in all frequency bands had little change before and after Fig. 4 EGG dynamic spectra of FD patients before
tr
eatment (a) and in a week treatment (b)
treatment (P>0.05), although FIC and PIC showed a e interval of each spectrum line is 1 min trend to decreasing (P>0.05). Fig. 5 IGM dynamic spectra of FD patients before (a), after one week (b) and three weeks (c) of treatment
The interval of each spectrum line is 1 min Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 Table 2 IGM parameters for 30 gastr
itis subjects and 30 healthy subjects
Control 68.5±26.5 60.1±3.3 1.36±0.08 0.18±0.16 Study 36.1±21.8 44.6±4.8 2.23±0.55 0.24±0.05
<0.05 <0.05 <0.05 <0.05 Data are expressed as mean±SD. PNF: percentage of normal frequency; PNP: percentage of normal power; FIC: frequency instability coef- ficient; PIC: power instability coefficient Table 3 EGG parameters for 30 gastritis subjects b efore and after one week of treatment
0–2 cpm Before 24.0±5.6 51.5±11.1 24.4±5.5 2.22±0.43 0.34±0.03
22.7±3.4 54.3±6.7 23.1±3.3 1.77±0.19 0.23±0.02
>0.05 >0.05 >0.05 <0.05 <0.05 Data are expressed as mean±SD. PP: percentage of power; FIC: frequency instability coefficient; PIC: power instability coefficient Table 4 IGM parameters for 30 gastritis subjects b efore and after one week of treatment
0–2 cpm Before 27.5±2.4 44.6±4.8 27.9±2.4 2.23±0.55 0.24±0.05
27.4±2.2 44.9±4.4 27.8±2.2 1.91±0.65 0.21±0.06
>0.05 >0.05 >0.05 >0.05 >0.05 Data are expressed as mean±SD. PP: percentage of power; FIC: frequency instability coefficient; PIC: power instability coefficient 4 Discussion
We also found a significant difference in IGM parameters between the gastritis subjects and the The results of this study indicate that the IGM of healthy controls. After one week of treatment, the FD subjects is not yet improved with one week of EGG rhythm of the gastritis subjects returned to treatment while the EGG is returned to normal. After normal, while some IGM parameters showed only a three weeks of treatment, the regular IGM rhythm of trend to improvement. This suggests that the influ- FD subjects is normal and shows a recovery of con- ence of gastric electric activity may not have coupled traction function of the stomach. The gastric motility with mechanical contraction of the stomach after only of human body is regulated by a series of sensory one week of treatment. From the clinical point of view, signals from the nervous system and has a close rela- although the subjects in the study had felt some alle- tionship with the electric activity of the gastric viation after a week of treatment, the symptoms of the smooth muscle (Zhou and Ke, 2005; Ma et al., 2006). gastritis were not completely relieved. The fact coin- While the EGG of FD subjects returned to normal cided with the results of Tables 3 and 4, and the after one week of treatment, this improvement may suggestion is that continued treatment should be not yet couple with or transfer to the mechanical ac- tivity of the stomach. After three weeks of treatment, EGG reflects the gastric electric activity of the when the influence of the electric activity has already stomach and is more sensitive to the regulation me- coupled with or transferred to mechanical activity of chanisms of nerve and electric activities. The im- the stomach via the regulation mechanism of nerve provement of the EGG after treatment is only the and electric activities, the normal IGM rhythm is seen beginning of the improvement of gastric motility in the spectra, and suggests the recovery of contrac- function, and it does not indicate the cure of gastric tion function of the stomach. disorder or the recovery of gastric motility. IGM is a Li et al. / J Zhejiang Univ-Sci B (Biomed & Biotechnol) 2011 12(12):983-989 veritable measure of gastric contraction and peristal- Physiol. Meas., 24(1):45-55. [doi:10.1088/0967-3334/
sis, and reflects the gastric motility function more Hadi, N.A., Giouvanoudi, A., Morton, R., Horton, P.W., effectively. Gastric study based on IGM measurement Spyrou, N.M., 2002. Variations in gastric emptying times and simultaneous EGG is a noninvasive, convenient, of three stomach regions for simple and complex meals and effective method. using scintigraphy. IEEE Trans. Nucl. Sci., 49(5):
2328-2331. [doi:10.1109/TNS.2002.803785] References
Huerta-Franco, R., Vargas-Luna, M., Hernandez, E., Capac- Chaw, C.S., Yazaki, E., Evans, D.F., 2001. The effect of pH cione, K., Cordova, T., 2009. Use of short-term change on the gastric emptying of liquids measured by bio-impedance for gastric motility assessment. Med. Eng. electrical impedance tomography and pH-sensitive ra- Phys., 31(7):770-774. [doi:10.1016/j.medengphy.2009.
diotelemetry capsule. Int. J. Pharm., 227(1-2):167-175.
Kothapalli, B., 1992. Origin of changes in the epigastric im- Chen, J.D., Lin, X.M., Abo, M., 2000. Multi-channel Gastric pedance signal as determined by a three-dimensional Electrical Stimulation for the Acceleration of Gastric model. IEEE Trans. Biomed. Eng., 39(10):1005-1010.
Emptying. First International Conference on Advances in [doi:10.1109/10.161332] Medical Signal and Information Processing. Bristol, UK, Li, Z.Y., Ren, C.S., 2008. Gastric motility measurement and evaluation of functional dyspepsia by a bio-impedance Chen, R.X., Wan, D.R., 1992a. Abnormal impedance gastro- method. Physiol. Meas., 29(6):S373-S382. [doi:10.1088/
gram and upper gastrointestinal symptom. Med. Biol. Eng. 0967-3334/29/6/S31] Li, Z.Y., Ren, C.S., 2009. Impedance gastric motility meas- Comput., 29(Suppl.):14-15.
urement to evaluate functional dyspepsia. Chin. J. Bio- Chen, R.X., Wan, D.R., 1992b. Further Investigation of Reli- med. Eng., 28(3):372-376 (in Chinese).
ability on Impedance Gastrograph for Continuous Meas- Li, Z.Y., Sha, H., Wang, Y., Zhao, S., Wang, W., Ren, C.S., urement of Human Gastric Contractile Activity. Pro- 2007. A New Approach of Gastric Motility Measurement ceedings of the 8th International Conference on Electrical and Evaluation by Bioimpedance. Proceedings of 13th Bio-impedance. July, Kupio, Finland, p.151-152. International Conference on Electrical Bio-Impedance & Deng, J., Zhao, S., Sha, H., Wang, Y., Ren, C.S., Li, Z.Y., 8th Conference on Electrical Impedance Tomography. 2008. Electrical Bioimpedance Method to Measure Gas- Aug. 29, Graz, Austria, p.691-694. tric Motility Function. Proceedings of the 2nd Interna- Li, Z.Y., Sha, H., Zhao, S., Wang, Y., Ren, C.S., 2008. tional Conference on Bioinformatics and Biomedical Liquid gastric emptying measurement using an electrical Engineering. Shanghai, China, p.616-619. [doi:10.1109/ bio-impedance method. Chin. J. Med. Instrum., 32(4):
Familoni, B.O., Kingma, Y.J., Bowes, K.L., 1987. Noninva- 253-256 (in Chinese). sive assessment of human gastric motor function. IEEE Liu, S.R., Li, Z.Y., Ren, C.S., Liu, C.L., Fang, X.J., 2009. Trans. Biomed. Eng., 34(1):30-36.
Research on gastric motility for erosive gastritis. Chin. J. [doi:10.1109/TBME. Med. Phys., 26(3):1224-1227 (in Chinese).
Gajre, S.S., Anand, S., Singh, U., Saxena, R.K., 2006. Novel Ma, D.S., Zhou, B., Fu, B., Song, M., 2006. Gastrointestinal Method of Using Dynamic Electrical Impedance Signals Motor and Clinical Aspects. Military Medicine Science for Noninvasive Diagnosis of Knee Osteoarthritis. Engi- Press, Beijing, p.37-44 (in Chinese). neering in Medicine and Biology Society. 28th Annual Ren, C.S., Li, Z.Y., Zhao, S., 2010. Use of electrical bioim- International Conference of the IEEE. New York, NY, pedance for gastric motility measurement and evaluation. p.2207-2210. [doi:10.1109/IEMBS.2006.260671] World Chin. J. Digestol., 18(1):1-8 (in Chinese).
Garay, L., Ramos, E.G., Cardiel, E., Munoz, R., Hernandez, Soulsby, C.T., Khela, M., Yazaki, E., Evans, D.F., Hennessy, P.R., 2006. In vivo and in situ measurement of electrical E., Jeremy, P.T., 2006. Measurements of gastric emptying impedance for determination of distention in proximal during continuous nasogastric infusion of liquid feed: stomach of rats. Med. Eng. Phys., 28(7):648-655.
electric impedance tomography versus gamma scinti- graphy. Clin. Nutr., 25(4):671-680. [doi:10.1016/j.clnu.
Giouvanoudi, A.C., Spyrou, N.M., 2008. Epigastric electrical impedance for the quantitative determination of the gas- Sutton, J.A., Thompson, S., 1985. Measurement of gastric tric acidity. Physiol. Meas., 29(11):1305-1317. [doi:10.
emptying rates by radioactive isotope scanning and epi- gastric impedance. The Lancet, 325(8434):898-900.
Giouvanoudi, A., Amaee, W.B., Sutton, J.A., Horton, P., Morton, R., Hall, W., Morgan, L., Freedman, M.R., Zhou, L., Ke, M.Y., 2005. Textbook of Neurogastroenterology Spyrou, N.M., 2003. Physiological interpretation of and Motility, Basic and Clinical Aspects. Science Press, electrical impedance epigastrography measurements. Beijing, p.501-521 (in Chinese).

Source: http://www.yiwey.net/jzus/oldversion/opentxt.php?doi=10.1631/jzus.B1000436