Vargas-Luna FM, Huerta-Franco MR, Delgadillo-Holtfort I, Balleza-Ordaz M, Murillo-Torres RM. Correlation of electrogastrography and bioelectric impedance techniques for the gastric motility assessment Biomed. Eng.-Biomed. Tech. 2025.

Авторы: Vargas-Luna F.M. / Huerta-Franco M.-R. / Delgadillo-Holtfort I. / Balleza-Ordaz J.M.  / Murillo-Torres R.M.

https://doi.org/10.1515/bmt-2024-0438

Correlation of electrogastrography and bioelectric impedance techniques for the gastric motility assessment

Abstract

Objectives

The electrical bioimpedance (EBI) technique has been used to measure gastric motility and emptying parameters. A well-known technique for this purpose is electrogastrography (EGG). No correlation between EGG signal and mechanical motility has been reported. In this study, a direct data comparison of these two techniques was performed.

Methods

23 volunteers underwent simultaneous gastric monitoring using EGG and EBI. Signal processing was performed to isolate the slow waves of 0.5–9 cpm. The parameters obtained from 70 % overlapped time slots of 3.5 min, included the dominant frequency and power of the normo-gastric region and the percentage of brady-, normo-, and tachy-gastric slow waves.

Results

The EGG showed slightly higher values in dominant frequencies, whereas EBI displayed higher variability. High-frequency features were more significant in the EBI, with lower variability, and correlations were found in approximately half of the frequency spectra. Slow waves exhibited poor correlation, but were significant at 95 % of the timeslots.

Conclusions

Comparing EBI and EGG, global parameters in the normogastric region had slight variances, which may not significantly impact clinical findings. The sensitivity of the EBI to higher frequencies is evident.

Keywords: bio-impedance; electrogastrography; gastric motility

Introduction

The recording of electrical activity of the stomach has been studied for decades. Alvarez WC [1] reported the “action currents in stomach and intestine”, and Davis, Garafolo & Gault [2] searched for the “potentials of the abdominal region.” It was not until 1975 [3] that electrogastrography (EGG) was considered a promising alternative to obtain information about gastric activity. The performance and clinical utility of EGG seems to be well established since 2003 [4], and later when Yin & Chen [5] reviewed the methodology, validation, and applications of this technique in general, and its validation using multichannel option and its use for children, by Riezzo, Russo & Indrio [6] in the same year.

One of the first proposals to search for gastric motility parameters by electrical bio-impedance (EBI) was introduced by Sutton, Thompson, and Sobnack [7], who compared the changes in time of the electrical impedance of the epigastric region with scintigraphy results about emptying when the subjects under evaluation ingested a low electrically conductive liquid. In the same year, this research group compared the EBI technique with the dye dilution method and used metoclopramide to fasten emptying [8]. In both cases, the excellent concordance between the EBI and gold methods for empty assessments suggests that detecting gastric contractile activity may also be possible. Despite these promising early results that foreshadow the utility of EBI for motility information, electrical activity and mechanical motility do not seem to easily attune each other [9].

The comparison of EBI and EGG techniques was addressed almost simultaneously as EBI was proposed for gastric evaluation. In 1991, Clevers, Smout, van der Schee, and Akkermans [10], assessing motility in patients with a postoperative condition (cholecystectomy and major colonic surgery), studied the difference between the results of the two aspects of gastric physiology: myoelectrical activity obtained by EGG and motor activity obtained by EBI; in fact, few direct comparisons have been made between both techniques. Smout, Jebbink, Akkermans, and Bruijs [11] asserted that EGG provides little information about gastric motility and emptying, and they compared EBI results with electrical tomography to declare their doubts about its clinical usefulness, but only in emptying. The group of Zhao et al. [12] has performed a more systematic comparison between both techniques, focusing on the application in specific pathological conditions such as functional dyspepsia, where they still emphasize the role of EGG to monitor electrical activity and EBI for gastric motility but looking at similar results in the frequency domain. In another study, the same group [13] studied patients with functional dyspepsia and showed differences between techniques in evaluating normalization of gastric condition after treatment.

The imaging option (Electrical Impedance Tomography or EIT) has been searched for good emptying parameters mainly compared with scintigraphy methodologies [7], 14], in contrast with other studies, where a direct look at the impedance change over time was considered to detect gastric emptying after food consumption.

Our group has worked on the EBI technique, using an overlapped short-term recording of 3.5 min [15], [16], [17] to simulate continuous parameter acquisition. Parameters in the frequency domain were considered in other situations (food consumption, medication to enhance motility and psychological stress).

Despite the similarities and differences between the two techniques, the parameters and methodologies used to analyze the data are the same.

Based on these previous studies, we should realize that EGG and EBI evaluate two aspects of the gastric condition: electrical activity and motility. EGG in fact correlates with gastric symptoms [18] and pathological conditions [19], 20]. Moreover, the physiological interpretation of EGG is well understood [21]. However, as mentioned before, EGG signals do not always correlate with EBI [9].

We hypothesized that even when both techniques measure different aspects of gastric activity and detailed results may differ, these differences are negligible, and the clinical consequences of these differences are not significant.

The objective is to quantify to what extent we can use the same parameters and methodologies interchangeably to analyze the EGG or EBI signal, that is, whether the EGG is suitable for measuring motility and giving the same clinical parameters as the EBI results, and use this information to draw conclusions about gastric performance.

Materials and methods

An analysis of datasets from 23 healthy subjects was performed (18–25 years old, 70 % men). All volunteers signed an informed consent form after they were informed about the objectives, procedures, possible risks, and benefits of the project in detail. A general protocol for the use of electrical bioimpedance to monitor gastric motility was approved by the Institutional Ethical Committee (CIBIUG-P-l0-2015). The inclusion criterion was to be healthy regarding gastrointestinal function. In this way, no volunteers had antecedents of diagnosed gastrointestinal diseases or symptoms, such as nausea, vomiting, blowing, or gastrointestinal distress in the week prior to the measurements.

To obtain information about gastric activity (electrical and mechanical), both EGG and EBI techniques were used.

EGG records the internal electrical activity of the body. In this case, an initial filtering (<1 Hz to cover some harmonics and >0.005 Hz) is important to discard the electrical activity of other muscles (mainly heart and musculoskeletal movements), and signal amplification is performed (from 100’s μV to at least tenths of volts, using a gain value of 1,000 or 2,000).

EBI measures the electrical impedance of the body region (gastric region). A small fixed electric current (400 μA) was injected at a frequency of 50 KHz using a pair of electrodes. This frequency does not interfere with the EGG recording and is the appropriate frequency for penetrating human tissue. The other two electrodes placed on the back recorded the voltage to calculate the impedance of the measured region. Again, a band pass filter and proper amplification similar to the case of EGG is used. In both cases, the amplification rate depended on the body composition of the volunteers. The electrodes type used are the same for both techniques and both signals are recorded simultaneously at 250 samples per second (s/s).

The data for both methodologies were processed identically (same decimated ratio, same filters and/or smoothing parameters, and the same FFT parameters). No interference is expected between techniques because EBI uses an excitation current with a frequency of 50 KHz, which is far from the filter limits. In general, the signals were standardized before processing to enable a similar analysis.

The electrode placements were chosen to coincide with the external anatomy and the internal position of the stomach in each subject, independent of the body build.

Four electrodes were placed for EBI recordings: one in the middle path between the xiphoid process and umbilicus, and the second in the midsternal on the subject’s left side, just below the lower rib and above the level of the first electrode. The other two electrodes were placed in the back at the same level as the frontal electrodes, avoiding the spinal cord. Two EGG electrodes were placed adjacent to the frontal EBI electrodes right sided, and the third one was in the right side of the hip.

The position of the electrodes is indicated mainly by anatomical references; therefore, distances and geometry in general could differ among individuals. In addition, BMI and other anatomical features influence the signal in magnitude but not in shape and frequency. As mentioned above, signal processing makes use of a standardized signal. This is needed because they are two signals coming from different sources and different physiological phenomena, so standardized signals make it possible to compare.

The protocol consisted of recordings registered simultaneously using BIOPAC equipment with modules for EGG and Bio-impedance at 250 (s/s) rates. The volunteers had an initial adaptation period of 5 min with all equipment and electrode wiring connected. After 5 min of adaptation, the volunteer remained at rest for 30 min in a supine position in a quiet, private room. The setup is shown schematically in Figure 1.

Figure 1: Scheme of the set up for data acquisition. Grey electrodes correspond to EBI wiring and black electrodes for EGG wiring.


Volunteers were asked to remain still as much as possible during the signal recordings to avoid mechanical or electrical artifacts arising from movements (abnormal respirations due to sighs, cough, sneeze, talk, or muscular contractions even from the upper or lower limbs). In general, these types of artifacts affect a wide range of frequencies. These events (voluntary or involuntary) were identified and registered during the experiments and later compared with the signal file. When these artifacts appear and remain visible after filtering, the time slot is discarded from the analysis.

Gastric motility consists of slow waves (resulting from a pacesetter potential) and gastric contractions (the counterpart of the spike potentials). Normal slow waves are around three cycles per minute (cpm) in humans (the normal range is from 2 to 4 cpm), lower frequency waves are considered Bradygastria (0.5–2 cpm), and higher frequency waves (4–9 cpm) are classified as Tachygastria [5]. The percentage of time or number of slow waves in each of these regions was calculated.

Data processing

The basic data-reducing and cleaning procedure consisted of decimating the data to 5 s/s, detrending, and smoothing using two-order polynomial fitting every 1-s interval. Artifact removal was performed using the “Linear Minimum Mean Square Error Estimation” (LMMSE) [22] and a further third-order Butterworth filter to have frequencies between 0.5 and 9 cpm.

Many methods have been used to obtain the frequency domain spectrum of gastric activity. We observed no significant differences in data processing using different methods. For comparison, we chose the Yule–Walker method because it is an autoregressive method used for stationary signals. In general EGG is considered non-stationary signal but as we are analyzing short time EGG signals, it is expected to maintain the statistical properties in that time period and this autoregressive method is appropriate [23].

To simulate the continuous monitoring of parameters obtained from time intervals (vg medians, FFT parameters), we took small time intervals that highly overlapped (70 %). The minimum time interval we consider that we can analyze is when we have at least 1.75 waves in the lowest bradygasric frequency (0.5 cpm) that is 3.5 min. In this case, FFT can give frequencies as low as 0.28 cpm, which is advantageous considering the soft tail of the low-frequency limit cut off (0.5 cpm) using an order 3 Butterworth filter.

Therefore, for the frequency domain parameter analysis, timeslots of 3.5 min with an overlap of 2.5 min were analyzed. The median of the corresponding parameters from the time slot results was analyzed for each parameter to be considered.

The parameters for comparison were the dominant peak and power in the normal region of the FFT spectrum if they existed, and the proportion of the area under the frequency spectrum in each region. Variability of each parameter is also analyzed.

Statistical analysis

Paired Wilcoxon signed-range tests were used because many parameters did not show a normal distribution. Comparison of median values, the measure of their standardized dispersion (variation coefficient: VC=standard deviation/average), and the correlation between the parameters for both techniques are reported. The analysis was performed using a MatLab platform.

Results

In Figure 2, we include an example of a time-domain comparison of signals in each activity region. In a typical time-domain signal, in both techniques, the waves are mixed in frequency (there exists gastric rate variability). For classification purposes, a normal dominant behavior is considered if more than 70 % of the slow-wave frequencies are in the corresponding normogastric region [5]. We do not classify gastric activity but only compare parameter values in the time or frequency domain.





In addition to the fact that the parameters obtained did not follow a normal distribution, the median of each parameter was the most reliable in this analysis to avoid the influence of abnormal parameter values due to remnant artifacts or noise after filtering. A comparison of the variation coefficient of each parameter provides an idea of the natural variability of the parameter. Therefore, these two parameters were considered in the results (Table 1), and are listed below.

• Dominant frequency (frequency at which the signal’s frequency spectrum has the higher peak) in the normogastric region (Fn) tends to be lower on average using EBI-technique (3.06 cpm compared with 3.17 of EGG) without reaching statistical significance (p=0.07, Figure 3a).


• Still, EBI’s variability (variation coefficient: VC) of dominant frequency in the normogastric region is significantly higher (0.22 vs. 0.13 of EGG, p=0.02, Figure 3b).


• Dominant power in the normogastric region is statistically equal in both techniques (p=0.13)


• Regarding the dominant power parameter, in normogastric region, there tends to be more variability using EBI than EGG (0.59 for EGG and 0.69 for EBI, p=0.055).

FFT distributions showed the influence of high-frequency noise, possibly mainly due to other gastrointestinal regions (bowels), respiration (participants having less than 10 breaths per minute), etc.

• The relative area of the high-frequency region in FFT (tachygastric region) is larger in EBI than using EGG (0.5 for EBI and 0.43 for EGG, p=0.02, Figure 4a).


• In consequence, EBI also shows a lower relative percentage of normal region activity than EGG (0.38 for EBI and 0.45 for EGG, p=0.01, Figure 4a).


• The variation coefficient for the relative area in tachygastric region is lower for the EGG technique (0.25 for EBI and 0.23 for EGG, p=0.045, Figure 4b).

Normo-, brady-, and tachy-gastric activities were also estimated by counting individual slow waves in each region (Table 2) observing the following facts.

• The normogastric slow waves were fewer in EBI compared with EGG (being 68 % for the EBI technique and 73 % for EGG, p=0.045, Figure 5a).


• Tachygastric slow waves tend to be larger in EBI compared with EGG. The increased high-frequency activity did not reach significance in this case (31 % for the EBI technique and 27 % for EGG, p=0.07, Figure 5b).


• Bradygastric activity was small and similar in both methods (0.35 % for EGG and 0.34 % for EBI, p=0.6).

If the percentage of time spent in each activity region is considered as an alternative to slow-wave counting, there was no statistical difference in any region for both techniques.

The correlation of the parameters obtained from short-time analysis (timeslots of 3.5 min overlapping 2.5 min) is, in general, very poor or not significant. However, some results were observed (Table 3).

• The frequency spectra in both techniques correlate significantly in 45 % of the timeslots, with an average correlation of R=0.55 (Table 3).


• In the time domain, if the slow waves are restricted to a near-normal region (1–5 cpm), the correlation between the signals is poor (R=0.22 on average) but significant in 96 % of the cases (Table 3).


• Nevertheless, the frequency calculated from the distance in time between adjacent slow waves correlated in 57 % of the timeslots, with an average R=0.42.

Discussion

Theoretically, the frequency resolution for 3.5-minutes-long data is approximately 0.3 cpm. Here, the data were processed to have 0.5 cpm up to 9 cpm waves, but slow wave analysis shortened the bradygastric region to 1 cpm. The data from both techniques were obtained simultaneously and processed in the same way; therefore, it was expected that the results, except for remnant noise and artifacts, would be the same for both techniques.

If at least two or more waveforms are analyzed, the frequency contribution in the FFT analysis is better defined. Therefore, bradygastria (0.5–2 cpm) was considered in the accuracy limit for this experiment (using 3.5-min timeslots). An extensive background at a low frequency (<1 cpm) appears in the FFT for both techniques, even when a detrending procedure is performed. This, together with the filter’s low-frequency cutoff that coincides with the bradygastric low boundary, yields fake bradygastric peaks that do not permit complete trust in the FFT results for that region. Nevertheless, these low-frequency features were similar for both techniques. The area under the frequency spectrum corresponding to the bradygastric regime, although relatively low (6–7 % of the total area on average), is one order of magnitude higher than the percentage of slow waves found in a more restricted range of 1–2 cpm (0.34–0.35 % on average). In these cases, the percentages were statistically similar for both techniques. This demonstrates the importance of low-frequency artifacts, the necessity of carefully detrending the data, and the fact that real bradygastric events are better defined with direct waveform counting.

On the other hand, tachygastria considered from 4 to 9 cpm overlaps with the motility of the small intestine and colon that both techniques should detect. EBI is more influenced by these high-frequency contributions than EGG. In the EGG technique, the electrodes are placed in such a way to maximize the detection of the depolarization in the stomach that yields to the corresponding muscle contraction. In the small intestine, these depolarization events occur in different spatial and temporal patterns, and some are weakly detected by the EGG electrodes. However, regardless of the origin of the movements, if motility is produced, the EBI technique should be capable of detecting it, because movement implies a change in the impedance of the region.

Therefore, EBI technique presents higher variability in dominant frequency and power and the relative area of tachygastric region. This greater dispersion ​​would be understood by the presence the noise sources that influence the EBI technique more than the EGG.

In the case of slow breathing, this external movement mainly affects the EBI technique because it does not involve electrical activity of the abdominal muscles. Although this contribution is minimized by asking the volunteer to breathe normally, resting tends to slow breathing, increasing activity in this high-frequency region for EBI. The increment in the relatively high-frequency spectra in the EBI was accompanied by a corresponding relative decrement in the normogastric region.

These differences in tachygastric contribution are observed again mainly in FFT spectra, but slow-wave counting, although it has the same trend, does not reach statistical significance. Normogastric percentage differences remained significant in the wave counting results. Finally, if wave counting is performed in time, the three regimes are statistically equal for both techniques: p=0.9, p=0.08, p=0.15 for brady-, normo-, and tachy-gastric regime respectively (Table 2).

Even when the dominant frequency in the normo-gastric region is statistically equal for both techniques, the p=0.07 (Figure 3a) suggests a tendency to lower values in EBI. This could be explained by the fact that some electrical activity did not yield mechanical motility. The dominant power in the normogastric region remained statistically equal for both techniques (p=0.13) (Table 1).

Most of the slow wave patterns in timeslots have significant correlations, regardless of whether they have low R values, and most percentages of slow waves in either number or time are statistically similar, except for normal slow waves in number. These percentages are those used to define normal pathological conditions (normal activity means >70 % of slow waves in normal values), indicating that both techniques are roughly identical to evaluate the main clinical parameters of normo- and brady-gastric conditions. EGG, however, seems to be better for tachygastric determination.

It is worth noting that EBI signals are complex in nature. In our data, the phase is generally very similar in waveform to the magnitude and, therefore, was not reported; therefore, the findings based on the magnitude waveform would be fairly the same as those obtained from the phase signal.

Multiple motility issues affecting the EBI signal and multiple electrical activities in other GI regions for EEG signals yield interferences that could hide real motility waveforms in the area of interest, and these could be different for each technique.

Restricting voluntary or involuntary movements is difficult to achieve; it is uncomfortable for subjects and often produces stressful conditions that alter their motility [17]. However, sleepy conditions also affect motility [24] and do not ensure stillness. A reasonable approach to real gastric motility can be achieved using any technique, each with its own advantages. EBI monitors motility more directly, which makes it more sensitive to other movements. This is an essential high-frequency contribution. EGG monitors electrical activity but detects poorly not well-oriented dipoles or dipoles sets with destructive interference.

In summary, a detailed point-to-point comparison of the techniques is generally low correlated either in the time domain or in the frequency domain; however, global parameters are very similar, with a slight overestimation of the tachygastric activity using the EBI technique and the corresponding underestimation of the normogastric activity.

As we filter the signal to have frequencies only between 0.5 and 9 cpm, is evident the presence of noise from other parts of the gastric system as colon and small intestine (9–12 cpm) [5] in electrical (EGG) and motility (EBI) signal. In both cases, the distance between the sources of that noise is expected to diminish the intensity of the interference, and the random polarization patterns further reduce this noise in the case of EGG. The respiratory rate in healthy adults is above 11 cpm [5], 25], affecting, only in the case of a very slow respiration pattern, mainly the EBI tachygastric recordings. Clinically, these small differences, although statistically significant, seem to be negligible in validating the indistinct use of any of these techniques (EGG and EBI) to evaluate the performance of gastric motility, except for tachygastric motility, in which EGG appears to be more suitable.

Conclusions

As some authors have established, each technique has a clear purpose: EGG to obtain information on electrical activity, and EBI to obtain information on internal conformational changes. Despite their shortcomings and advantages searched and documented in many works previously published, and even with the evident lack of detailed correlation of most of the parameters due to their evident differences in the physical parameters they record, both techniques give roughly similar information for clinical purposes mainly in the case of normogastric behavior, if global parameters as mean normal peak position, and relative power of normo-gastric PSD frequency region are considered. The percentage of slow waves in number yields an underestimation for normal activity in the case of EBI but there are no differences for other regions or percentages in time. Low-frequency noise affects both techniques similarly, nevertheless, the EBI technique is more sensitive to high frequency contributions (other GI motility or slow respiration patterns) than EGG, limiting its use in the case of tachygastric evaluation.

*Corresponding author: Francisco M. Vargas-Luna, Departamento de Ingeniería Física, Universidad de Guanajuato, Loma del Bosque 103, Lomas del Campestre, León, Gto., CP 37150, México, E-mail: francisco.vargas@ugto.mx
Maria-Raquel Huerta-Franco, Departamento de Ciencias Aplicadas Al Trabajo, Universidad de Guanajuato., León, Gto., México, E-mail: huertafranco@hotmail.com. https://orcid.org/0000-0001-7935-5151
Isabel Delgadillo-Holtfort and Marco Balleza-Ordaz, Departamento de Ingeniería Física, Universidad de Guanajuato, León, Gto., México, E-mail: idelgadilloh@ugto.mx (I. Delgadillo-Holtfort), jm.balleza@ugto.mx (M. Balleza-Ordaz). https://orcid.org/0000-0003-1961-0952 (I. Delgadillo Holtfort). https://orcid.org/0000-0002-3246-0277 (M. Balleza-Ordaz)
Regina M.Murillo-Torres, Escuela de Medicina y Ciencias de La Salud Del Tecnológico de Monterrey, Monterrey, N.L., México, E-mail: regina_murillo@hotmail.com. https://orcid.org/0009-0001-6245-7474

Funding source: Dirección de Apoyo a la Investigación y al Posgrado, Universidad de Guanajuato

Award Identifier / Grant number: 224/2019

Acknowledgments

This work was supported by DAIP (Research and Postgraduate Support Department), University of Guanajuato Mexico. Authors also thank all volunteers for kindly accept to participate in this research.

Research ethics: A project related to the use of Electrical Bio-impedance to monitor gastric motility was approved by the Institutional Ethical Committee of the University of Guanajuato ref: CIBIUG-P-10-2015. The approved protocol included detailed description of the subjects management and the national and international regulations observed during the study mainly the Declaration of Helsinki (2013).

Informed consent: All volunteers signed an informed consent, after they were informed in detail about the objective, procedures, possible risks and benefits of the project.

Author contributions: Francisco M. Vargas-Luna* and Raquel Huerta-Franco were the technical and medical responsible of the protocol, the data acquisition, processing and analysis. They also write the manuscript. Francisco M. Vargas-Luna and Raquel Huerta-Franco contributed equally to this work and share first authorship. Isabel Delgadillo-Holtfort and Marco Balleza-Ordaz help substantially in data processing and statistical analysis. Regina M. Murillo-Torres made the revision of the manuscript mainly in the bibliography, antecedents, interpretation of the results and discussion. The authors have accepted responsibility for the entire content of this manuscript and approved its submission.

Use of Large Language Models, AI and Machine Learning Tools: None declared.

Conflict of interest: The authors declare that there is no financial and/or personal interest or belief that could affect the objectivity of this work.

Research funding: This work was supported by DAIP (Research and Postgraduate Support Department), University of Guanajuato Mexico.

Data availability: The raw data can be obtained on request from the corresponding author.

References

1. Alvarez WC. The electrogastrogram and what it shows. JAMA 1922;78: 1116–18.
2. Davis RC, Garafolo L, Gault FP. An exploration of abdominal potentials. J Comp Physiol Psychol 1957;50:519–23.
3. Brown BH, Smallwood RH, Duthie HL, Stoddard CJ. Intestinal smooth muscle electrical potentials recorded from surface electrodes.MedBiol Eng 1975;13:97–103.
4. Parkman HP, Hasler WL, Barnett JL, Eaker EY. Electrogastrography: a document prepared by the gastric section of the American motility society clinical GI motility testing task force. Neuro Gastroenterol Motil 2003;15:89–102.
5. Yin J, Chen JD. Electrogastrography: methodology, validation and applications. J Neurogastroenterol Motil 2013;19:5–17.
6. Riezzo G, Russo F, Indrio F. Electrogastrography in adults and children: the strength, pitfalls, and clinical significance of the cutaneous recording of the gastric electrical activity. BioMed Res Int 2013;2013:282757.
7. Sutton JA, Thompson S, Sobnack R. Measurement of gastric emptying rates by radioactive isotope scanning and epigastric impedance. Lancet 1985;1:898–900.
8. Mc Clelland GR, Sutton JA. Epigastric impedance: a noninvasive method for the assessment of gastric emptying and motility. Gut 1985;26:607–14.
9. Abid S, Lindberg G. Electrogastrography: poor correlation with antroduodenal manometry and doubt fulclinical use fulness in adults.World J Gastroenterol 2007;13:5101–7.
10. Clevers GJ, Smout AJ, Van der Schee EJ, Akkermans LM. Myo-electrical and motor activity of the stomach in the first days after abdominal surgery: evaluation by electrogastrography and impedance gastrography. J Gastroenterol Hepatol 1991;6:253–9.
11. Smout AJ, Jebbink HJ, Akkermans LM, Bruijs PP. Role of electrogastrography and gastric impedance measurements in evaluation of gastric emptying and motility. Dig Dis Sci 1994;39:110S–113S.
12. Zhao S, Sha H, Li ZY, Ren C. Electrical bio-impedance gastric motility measurement based on an electrical-mechanical composite mechanism. World J Gastroenterol 2012;18:3282–7.
13. Li Z, Ren C. Gastric motility measurement and evaluation of functional dyspepsia by a bio-impedance method. Physiol Meas 2008;29:S373–82.
14. Soulsby CT, Khela M, Yazaki E, Evans DF, Hennessy E, Powell-Tuck J. Measurements of gastric emptying during continuous nasogastric infusion of liquid feed: electric impedance tomography versus gamma scintigraphy. Clin Nutr 2006;25:671–80.
15. Huerta-Franco R, Vargas-Luna M, Hernandez E, Capaccione K, Cordova T. Use of short-term bio-impedance for gastric motility assessment. Med Eng Phys 2009;31:770–4.
16. Huerta-Franco R, Vargas-Luna M, Hernandez E, Capaccione KM, Yañez Roldán E, Hernández-Ledezma U, et al. Effects of metoclopramide on gastric motility measured by short-term bio-impedance. World J Gastroenterol 2009;15:4763–9.
17. Huerta-Franco R, Vargas-Luna FM, Montes-Frausto JB, Morales-Mata I, Ramirez-Padilla L. Effect of psychological stress on gastric motility assessed by electrical bio-impedance. World J Gastroenterol 2012;18: 5027–33.
    
18. Gharibans AA, Coleman TP, Mousa H, Kunkel DC. Spatial patterns from high-resolution electrogastrography correlate with severity of symptoms in patients with functional dyspepsia and gastroparesis. Clin Gastroenterol Hepatol 2019;17:2668–77.
19. Kayar Y, Danalioglu A, Kafee AA, Okkesim S, Sentürk H. Gastric myoelectrical activity abnormalities of electrogastrography in patients with functional dyspepsia. Turk J Gastroenterol 2016;27: 415–20.
20. Koch KL. Electrogastrography for suspected gastroparesis. In: McCallum RW, Parkman HP, editors. Gastroparesis: Academic Press; 2021:189–205 pp.
21. Giouvanoudi A, Amaee WB, Sutton JA, Horton P, Morton R, Hall W, et al. Physiological interpretation of electrical impedance epigastrography measurements. Physiol Meas 2003;24:45–55.
22. Gharibans AA, Smarr BL, Kunkel DC, Kriegsfeld LJ, Mousa HM, Coleman TP. Artifact rejection methodology enables continuous, noninvasive measurement of gastric myoelectric activity in ambulatory subjects. Sci Rep 2018;8:5019.
23. Miljković N, Milenić N, Popović NB, Sodnik J. Data augmentation for generating synthetic electrogastrogram time series. Med Biol Eng Comput 2024;62:2879–91.
24. Wang G, Chen Y, Liu H, Yu X, Han Y, Wang W, et al. Differences in intestinal motility during different sleep stages based on long-term bowel sounds. Biomed Eng Online 2023;22:105.
25. Toften S, Kjellstadli JT, Thu OKF, Ellingsen OJ. Noncontact longitudinal respiratory rate measurements in healthy adults using radar-based sleep monitor (somnofy): validation study. JMIR Biomed Eng 2022;7: e36618.

This work is licensed under the Creative Commons Attribution 4.0 International License.

Supplementary Material

Назад в раздел