Materials and Methods

Animals—Six healthy dogs of various hound breeds were used in this blinded, prospective study. The study was approved by the University of Guelph Animal Care Committee. Dogs were 2 to 3 years old and weighed 20.2 to 26.1 kg (mean weight, 22.7 kg). For each dog, physical examination revealed no abnormalities; there was no history of gastrointestinal tract disease, and anesthetic agents or gastric motility–modifying medications were not administered within the 4-week period prior to commencement of the study.

Experimental phases—The experiment was performed in 3 phases. First, food was withheld from the dogs for 12 hours (food-withholding phase); during this period, abdominal ultrasonography and EGG were performed. The dogs were then fed 250 g of a complete dry food ration; after 30 minutes (postprandial phase), abdominal ultrasonography and EGG were performed. Finally, 60 minutes after feeding, atropine (0.04 mg/kg, IM) was administered; after 30 minutes (ileus phase), abdominal ultrasonography and EGG were performed. The person (HD) performing the ultrasonographic measurements was unaware of the experimental phase.

EGG recordings—Each dog was placed in dorsal recumbency, and the area over the stomach was clipped and shaved; in the area of electrode placement, the skin was abraded with alcohol-soaked gauze squares. As previously described,^12,14,15 5 standard neonatal ECG surface electrodes^b were positioned with coupling gel along the abdominal projection of the stomach axis (Figure 1). The correct position of the electrodes was evaluated ultrasonographically. During the EGG acquisition, an 8-channel bipolar recording system was used, which allowed recording of short- (channels 1 to 4) and long- (channels 5 to 8) distance channels (Appendix). During each 30-minute recording session, dogs were maintained in dorsal recumbency and motion of the dog was recorded to allow removal of motion artifacts from the raw EGG signal before analysis. Electrogastrographic signals were filtered in the frequency band of 0.02 to 0.15 Hz. The signals were amplified and digitized with a 10-Hz sampling frequency by use of a 16-channel analog-to-digital converter^c that was controlled by a personal computer. For each channel, 3- dimensional plots in the range of 1.8 to 12 cpm were obtained by use of the fast Hartley transform.¹⁶ The dominant spectral peaks in these plots were connected with lines to form 2-dimensional time-frequency plots, and the probability density function of the frequencies present in each channel were determined.¹⁷ The mean cpm value, variance, and SD of the dominant frequency obtained from the time-frequency plots of the different EGG channels were calculated to evaluate the stability of the dominant spectral component. The EGG recording was considered stable if the SD of the dominant frequency was < 15% in at least 3 of 8 EGG channels.^a The absolute value of the amplitude (in MV) of the EGG signal was calculated for each second; these values were each squared averaged over the 30-minute recording session to provide the total power (MV²). Treatment effect (postprandial or ileus phase) was expressed as the difference in total power between the treatment and food-withholding phases or power ratio (ie, the ratio of the power of the treatment phase measurement divided by the power of the baseline [food-withholding phase] measurement).^{9, 18}

Schema of the placements of 5 electrodes along the ventral abdominal projection of the stomach axis in a dog and the combinations forming the 8 bipolar channels used for multichannel EGG. The electrodes were designated A through E.

Citation: American Journal of Veterinary Research 70, 1; 10.2460/ajvr.70.1.11

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Schema of the placements of 5 electrodes along the ventral abdominal projection of the stomach axis in a dog and the combinations forming the 8 bipolar channels used for multichannel EGG. The electrodes were designated A through E.

Citation: American Journal of Veterinary Research 70, 1; 10.2460/ajvr.70.1.11

• Download Figure
• Download figure as PowerPoint slide

Schema of the placements of 5 electrodes along the ventral abdominal projection of the stomach axis in a dog and the combinations forming the 8 bipolar channels used for multichannel EGG. The electrodes were designated A through E.

Citation: American Journal of Veterinary Research 70, 1; 10.2460/ajvr.70.1.11

• Download Figure
• Download figure as PowerPoint slide

B-mode ultrasonography—B-mode ultrasonography was used as previously described¹⁹ to evaluate motility of the gastric antrum. The contractions of the gastric antrum were counted during each minute of a 5-minute period by an investigator (HD) who was unaware of the experimental phase; a contraction rate (mean number of contractions per minute) was calculated.

Statistical analysis—A 2-way ANOVA for a mixed model was performed on the collected data (cpm values, power ratios, and contraction rates). Their interactions were modeled to determine whether there were any significant differences among experimental phases for any of the 8 EGG channels or the ultrasonographic measurements. A Shapiro-Wilk test was used to confirm whether the data were normally distributed. Log arithmic transformation was applied where appropriate to meet the assumptions of normality. A Tukey comparison of means was performed if the F-test finding was significant. A multivariate t-adjustment was used to determine whether power ratio for the postprandial or ileus phase was significantly different from a value of 1 (food- withholding phase), and a post hoc Tukey test was used to compare power ratios between experimental phases. A Pearson correlation was performed to determine correlations among cpm value, power ratio, and contraction rate. Significance was set at a value of P b 0.05, and all analyses were performed by use of computer software.^d

Results

B-mode ultrasonography—In the food-withholding phase, mean contraction rate in the study dogs was 0.5 contractions/min (range, 0 to 1 contractions/min). In the postprandial phase, mean contraction rate was 4.1 contractions/min (range, 2 to 5 contractions/min). In the ileus phase, contractions were completely absent. Contraction rate differed significantly (P < 0.001) among phases; significantly fewer contractions per minute were detected in both the ileus (P = 0.001) and food-withholding (P = 0.02) phases, compared with findings in the postprandial phase. There was no significant difference in contraction rate between the food- withholding and ileus phases.

EGG data—The mean cpm values and SDs of the dominant frequency obtained from the time-frequency plots were calculated (Table 1). The criterion for stability was met in 4 of 6 dogs but only in the long-distance channels. The dominant frequency was most stable in the ileus phase. For any channel, no significant differences among the 3 phases were detected.

Mean power ratios (ie, power measured in the treatment phase divided by the power measured in the food-withholding phase) and SE values were calculated (Table 2). Compared with the ileus phase, the power ratio was significantly higher in the postprandial phase. For all 8 channels, a significant (P < 0.001) and good correlation between EGG power and ultrasonographic findings (contractions per minute) was evident during the postprandial and ileus phases. The Pearson correlation r values were as follows: channel 1, 0.747; channel 2, 0.789; channel 3, 0.691; channel 4, 0.716; channel 5, 0.727; channel 6, 0.722; channel 7, 0.633; and channel 8, 0.777.

Discussion

The results of the present study indicated that EGG can be performed in nonanesthetized dogs; the criterion for stability (ie, SD of the dominant frequency determined via time-frequency plots < 15% in at least 3 of the 8 channels¹⁴,^a) was met in most of the evaluated dogs. In another study²⁰ in which multichannel EGG was used to evaluate cecal and colonic myoelectrical activity in awake horses, the criterion for stability was not met. Stability is an important consideration in EGG because the reliability of dominant frequency assessment is dependent on stability.¹⁰ On the basis of data obtained in experiments in dogs and in human volunteers, it is known that if the criterion for stability is not met, then it is not possible to truly determine which frequency is dominant. In humans, the dominant frequency is thought to reflect the frequency of the slow waves and is associated with the actual frequency of gastric contractions.²¹ More recently, it was determined that the dominant frequency was not useful for evaluation of changes in myoelectrical activity in humans with naturally occurring gastric motility disorders.²² In the dogs of the present study, the dominant frequency in the different phases of the experiment did not differ significantly, even though the number of stomach contractions changed depending on the experimental phase. Thus, it appears that the dominant frequency is also not associated with the actual frequency of gastric contractions in dogs.

For noninvasive assessment of gastric myoelectrical activity, a well-established alternative to evaluation of the dominant frequencies is to assess total power^{10, 23}; in addition to its use in the study reported here, assessment of total power has been used in humans to evaluate the effect of erythromycin on gastric motility.¹⁸ The thickness of the body wall, position of and distance between electrodes, and method of spectral analysis all have an influence on the total power; therefore, it is recommended to clinically evaluate total power only in relation to a baseline measurement within the same individual.¹⁰ However, the calculation of a power ratio accounts for some of these factors.^9,10,18 The clinical interpretation of total power values in humans is somewhat controversial. Total power is considered by some to be directly related to the strength of intestinal contractions,^{10, 24} whereas it is considered by others to represent an assessment of myoelectrical activity but not intestinal contractions.²⁵ Compared with the evaluation of total frequencies, increased power ratios appear to be the only EGG variable that is repeatedly correlated with delayed gastric emptying in humans.²² In the dogs of the present study, the mean power ratio in all 8 channels was higher in the postprandial phase, compared with the value in the ileus phase. The EGG findings were also in agreement (ie, significantly and positively correlated) with the ultrasonographic findings; the mean number of contractions per minute in the postprandial phase was significantly higher than the mean number of contractions per minute in the ileus phase. It would have been interesting to replicate the experiments in the dogs 2 or 3 times to see whether the increase in power ratios in the postprandial phase, compared with findings in the ileus phase, would have persisted as investigated in a study¹⁰ in humans who were fed test meals after which power ratios were calculated repeatedly to ensure consistency of EGG.

To our knowledge, this is the first report of 8-channel EGG performed in healthy, nonanesthetized dogs; however, other researchers have studied the technique in a variety of situations. Atanassova et al²⁶ compared the electrical activity of the gastric muscle wall by use of electrogastromyography (via chronically implanted electrodes on the muscle wall of the stomach) and single-channel EGG in 5 mongrel dogs. Results of that study indicated that the slow-wave amplitude in the EGG recording increased after feeding, which corresponded to bursts or groups of spike potentials in the electrogastromyogram.

In a more recent study,²⁷ the use of 3-channel EGG to measure coupling or uncoupling of gastric slow waves was evaluated in healthy, nonanesthetized dogs. In 5 dogs, internal electrodes were implanted on the serosa of the stomach and gastric myoelectrical activity was measured in combination with EGG. Results indicated that 3-channel EGG was capable of detecting coupling and uncoupling of gastric slow waves.

Overall, use of multichannel EGG appears more desirable than use of single-channel EGG. In particular, information regarding the propagation and coordination of gastric slow waves cannot be derived via single- channel EGG.²⁷ Furthermore, the 8-channel EGG used in our study may offer some advantages over 3-channel EGG. For example, the increased number of leads (both short and long distance) ensures that the criterion for stability is met in nonanesthetized animals. Also, power ratios are available for 8 channels; if there is excessive variation among the 8 power ratios, this is evidence that the measurements should be repeated because the results may be erroneous. In conclusion, the study of this report was a valuable starting point for evaluation of the applicability of EGG in healthy, nonanesthetized dogs; results suggest that further studies (involving a larger number of patients) are warranted, particularly considering the potential clinical relevance of this technique.