User:John Andrew Sutton/sandbox

EPIGASTRIC IMPEDANCE

A non-invasive method for monitoring gastric function and the effect of treatment when the stomach is malfunctioning.

Principle of Epigastric Impedance Method.

The physical principle at the heart of Epigastric Impedance (EI) is that a small electrical current passing through the body meets a resistance (or impedance if the current is alternating) in each tissue that is directly proportional to its volume. Hence, when A/C is applied to the region of the stomach from electrodes on the skin of the abdomen and back, the gastric contents present an impedance proportional to their volume. So, when a subject drinks a liquid there is a marked change in overall impedance as shown in Figure 1. After the peak the rate at which the change returns to baseline has been shown to reflect gastric emptying.

Figure 1. The vertical axis represents the total impedance between electrodes on the skin of the upper abdomen and the back placed so the stomach lies between them. After 10 minutes of reasonably stable baseline there is an increase of about 3 ohm in body surface impedance just after the subject finishes swallowing a liquid test meal. The increase is directly proportional to the volume swallowed and the specific resistivity of the fluid in Ohms/cc-1.[1]

The device

To date no monitor is commercially available and past researchers have used prototypes made by their own electronic workshops. They are little more than a source of an alternating current, usually in the range 3-5 milliamps, coupled to a laptop computer that displays current flow and records the signal. For safety they are separated from the subject by a fibre-optic cable that transmits the signal and avoids direct connections between mains equipment and the subject. The laptop also hosts software to analyse the signal and present it in a helpful form.

Method

Suitable liquids have different impedances from body tissues because they change the impedance overall; in a positive direction if the liquid is more impedic than the baseline value and in a negative direction if the liquid is less so. Liquids that are iso-conductive (ie: the same as body tissues) produce no deflection.

The size of deflection is proportional to the volume and the specific impedance of the liquid so a useful standard test “meal” is de-ionised, pure water due to its high specific impedance approximately 100,000 times greater than body tissues[2] Resistivity can be further increased by adding non-ionised molecules such as glucose, and 5 or 10% solutions have been used previously.[3,4]

The subject must be fasting so that the stomach is empty. Otherwise conductive gastric contents can mask the fluid “meal” and prevent a deflection. Subjects must be comfortably seated because they must avoid moving the torso and abdomen for the duration of recording, typically a total of some 45 to 60 minutes. Reclining at 45 degrees is standard. The technician will record a baseline period before the subject drinks and it needs to be a constant as possible, particularly if the test includes a motility assessment.

Standard, silver-silver chloride electrodes are suitable. The first of 4 may be placed halfway down the margin of the ribs on the left and medially adjacent to it so that the current does not pass through the rib margin. The second of 4 is best placed next to it on the opposite side to the ribs. Electrodes 3 and 4 are placed on the back so that the body of the stomach lies between them and the first two. Some studies used a third pair in order to survey a wider region of the epigastrium; the 5th anterior electrode being placed a further 5 cms lower and more lateral than the 2nd

Limitations

The technique is limited to liquid gastric emptying rates because solid or semisolid foods invariably have similar impedances to body tissues so epigastric impedance cannot detect them. Complex liquid foods provoke secretions from the intestinal tract that are themselves conductive which distorts the gastric emptying result, so the technique uses simple, crystalline liquids to avoid them. An early experiment addressed this by sampling test meals of 750ml and showing that secretions did not appear until 12 minutes when the half-emptying time was approximately 5 minutes.[5]

Epigastric Impedance for Measuring Gastric Emptying

Gastric Emptying Studies in volunteers After the peak, that is reached soon after the subject has finished swallowing the liquid, there is a decline that shows how fast the liquid is leaving the stomach so it provides a possible way of measuring gastric emptying. This has been verified in several volunteer and patient trials. The rate is broadly exponential so it is usually expressed as time to 50% emptying of the remaining contents.

The first publication reported a significant correlation, (R2 = 0.73), between Epigastric Impedance and the “gold standard” method of scintigraphy in 6 volunteers.[6] In the second it also matched an established dye dilution method in 5 volunteers.[5]

Then reliability of epigastric impedance emptying results was tested in 3 volunteer trials where changes in gastric emptying were expected in order to find out whether epigastric impedance detected them. It did so on each occasion. The first trial compared emptying rates when 5 healthy volunteers drank either 750ml distilled water or distilled water containing 5% glucose.[5] The glucose doubled the 50% emptying time from 10 to 20 minutes (p<0.05). In that experiment the volunteers also took metoclopramide, a medicine that stimulates gastric contractions and it produced the predicted faster 50% emptying time than placebo (p<0.001). The second trial compared sitting upright at 90 degrees and semi-reclining at 45 degrees, expecting faster empting when fully upright due to greater hydraulic pressure on the gastric outlet. [7] When 7 volunteer students drank 600ml of diluted orange squash while sitting upright their group average time was 6.8 minutes, approximately half the result of 12.9 when they sat at a reclining angle of 45 degrees. The third test was to detect gastric delay induced by an opiate analgesic. When morphine 10 mg was given to 12 volunteers it produced an average 50% emptying rate of pure water of 21±9.0 minutes compared with placebo at 5.5±1.9.[8] Then, as predicted from the pharmacology, the opiate antagonist naloxone prevented the delay, shortening the half-time back to 7.3±3.0 minutes, which was not different from placebo. These results were confirmed by measuring the rate that paracetamol was absorbed, using areas under the plasma concentration curve to 90 minutes. It was also noted that gastric emptying rates correlated with the presence and intensity of nausea caused by morphine. A similar effect of the opiate analgesic meperidine was found by O’Sullivan et al who used epigastric impedance to measure gastric emptying in women who had taken it during labour.[9] The average 50% emptying for those who took meperidine was 18.2 ± 4.0 minutes and for the untreated controls it was 10.3 ±1.4 (p< 0.05).

Gastric Emptying Studies in patients

Epigastric impedance has been used in clinical conditions where delayed gastric emptying is known to occur: in migraine, diabetes and paediatric practice.

Boyle and her colleagues measured gastric emptying by epigastric impedance in 46 migraine patients outside an attack (average 10.1±5.3 minutes) and in 64 controls who did not have migraine (averages 8.7±4.9 to 12.3 ± 5.9 minutes).[10] During 20 attacks 13 patients reported headache pain worse than mild and all their 50% gastric emptying times were 29 minutes or more.(p<0.001) When headache pain was reported as mild in the other 7 attacks all but one produced 50% gastric emptying times in the normal range and they averaged 13.0 minutes. The exception was 29.0 minutes and was accompanied by nausea.

Gilbey and Watkins were the first to measure gastric emptying in patients attending a diabetic clinic who had autonomic neuropathy.[11] They measured gastric emptying by EI in 22 patients who produced a mean gastric emptying rate of 12.5 minutes with a range of 6 to 30 minutes. The wide range was thought to be due to some patients not having gastric involvement in their neuropathy. Their mean was significantly different from that in 15 healthy controls, which was 8 minutes with a range of 3-17 (p<0.01)

Boulal studied 20 healthy controls, 40 diabetic patients, 20 with type 1 and 20 with type 2 disease.[12] Respective averages and standard deviations were 9.4 ±3.4, 24.0 ±9.2, and 19.9 ±8.5 minutes and the two diabetic groups differed significantly from the healthy volunteers. A subgroup of 8 patients who had cardiac involvement in a severe neuropathy had a slower average at 32.1 minutes. This indicated that epigastric impedance could be useful in diabetes, possibly by early detection of slow gastric emptying so that it could be treated sooner.

Smith and her colleagues established that gastric emptying rates in children in their clinic in Birmingham UK were the same as adults. [13] In two newborn patients with dumping syndrome they recorded that gastric emptying was exceptionally rapid. They recorded that their normal practice of thickening the babies’ feed produced slower emptying

Epigastric Impedance for Measuring Gastric Motility

In Figure 1 the trace is a broad line because it includes waves due to respirations at about 14-20 cycles per minute. They are caused by a descending diaphragm compressing the stomach as the subject breathes in. It also includes slower waves at a frequency of 1 to 5 cycles per minute which corresponds to gastric contractions. These may be quantified by Fourier analysis to show the motile response to a meal.[5] Fourier transform is the normal method used to analyse waveforms emanating from the brain in electroencephalography (EEG), including assessments of the dominant frequency and power. For epigastric impedance Short Time Fourier Transform (STFT) is applied to epochs of approximately 5 minutes’ duration to produce a power spectral density (PSD) for each selected epoch and when PSDs are plotted in chronological order a 3-D plot of time versus frequency versus power is obtained as in Fig. 2. Resolution of the time axis may be improved by overlapping the epochs and in EI research epochs of five minutes are used with an overlap of 80%, giving time intervals of 1 minute.[3,14] Figure 2: STFT of the signal in Fig 1 using epochs of 5 minutes, overlapped by 80%.

Along the left margin frequencies in the range of 1.5 to 4.2 cycles per minute (cpm) represent gastric contractions. They form a biphasic line of peaks that begins only after 10 minutes of baseline recording. Frequencies on the right hand side of the valley represent respirations at 14 to 18 cpm. (From Giouvanoudi 2000[3])

In Fig 2 the power density in the 2.4 to 3.6 cpm frequency bandwidth, shown on the vertical axis, is absent in the 10-12 minutes before the meal but it appears as soon as the meal is ingested. Activity increases and remains high until about 35 minutes. The later burst around 45-55 minutes is likely to be a fasting motor complex because the volunteer would have been fasting and the test “meal” only water.

The response to a test meal is calculated from the ratio of power post vs pre meal. McClelland and Sutton first showed that prokinetic drugs increase this ratio.[5] The dominant frequency of the contractions in this frequency range is also calculated by taking the frequency with the highest power for all epochs. The result may be plotted against time to show how the dominant frequency of the contractions is moderated.

An academic group in China has defined normal contractility as contractions within the 2 – 4 cycles per minute waveband, expressing it as a percentage of all the contractility up to 5 cpm.[15] They called it percentage of normal frequency (PNF). They found that the normal subjects’ group PNF was 68% while in 28 dyspepsia patients it was only 28% (p< 0.01). They found similar figures for the patients who had erosive gastritis since their group PNF was 36% (p< 0.05). Similar calculations for power instead of frequency produced similar differences between normal and the two patient groups. This group also treated their functional dyspepsia patients with the prokinetic domperidone which reversed the symptoms after 3 weeks’ treatment. At the same time their emptying rate and their motility patterns normalized, showing that the epigastrograph could accurately record successful treatment.

Summary

Epigastric Impedance is a non-invasive method for measuring gastric emptying of simple, crystalline liquids. Initial studies appear to show it capable of providing accurate diagnosis of pathologically delayed emptying and motility.

References

1.	T. Fenlon. Medical Applications of Electrical Impedance Measurements. (1992) PhD Thesis, University of Surrey

2.	A. Iverson. The Measured Resistivity of Pure Water and Determination of the Limiting Mobility of OH- from 5 to 55° J Phys Chem (1964) 68 3 pp. 515-521

3.	Giouvanoudi A. Electrical Impedance Measurements In Gastric Function Investigations. PhD Thesis University of Surrey. 2000.

4.	N.A. Hadi. Impact of Different Calorific Meals and Pharmacological Blocker Agent on the Emptying Behaviour of the Whole Stomach and its Three Regions Using Simultaneous Scintigraphy and Electrical Impedance Epigastrography. (2004) PhD Thesis, University of Surrey

5.	G.R. McClelland, J.A. Sutton. Epigastric impedance; a non-invasive method for the assessment of gastric emptying and motility. Gut (1985), 26 6 pp. 607-614

6.	Sutton JA, Thompson S and Sobnack R. 1985. Measurement of gastric emptying by radioisotope scanning and epigastric impedance. Lancet i: 898-900.

7.	Rainbird AL, Pickworth MJW, Lightowler C, Mitchell M, Wingate DL.1987. Effect of posture and cold stress on impedance measurements of gastric emptying. Pharmaceutical Medicine (London) 2(1): 35-42

8.	Murphy DB, Sutton JA, Prescott LF, Murphy MB. 1997. Opioid induced delay in gastric emptying. A peripheral mechanism in humans', Anesthesiology, 87 765-770.

9.	O’Sullivan GM, Sutton JA, Thompson SA and Carrie LE. 1987. Non-invasive measurement of gastric emptying in obstetric patients. Anesth Analg, 66: 505-511.

10.	Boyle R, Behan PO and Sutton JA.(1990) A Correlation Between Delayed Gastric Emptying & Severity of Migraine Measured by the Epigastric Impedance Method. British J Clin Pharmacol 30: 405–9.

11.	Gilbey SG and Watkins PJ. 1987. Measurement by epigastric impedance of gastric emptying in diabetic autonomic neuropathy', Diabetic Medicine 4, 122-126

12.	Boulal IS. 2005 Evaluation of Gastric Emptying and Motility in Healthy and Diabetic Subjects by Electrical Impedance Epigastrography. PhD Thesis. University of Surrey May 2005.

13.	Smith HL, Newell SJ, Puntis JWL Hollins GW and Booth IW. Use of Epigastric Impedance Recording to measure Gastric Emptying in two Infants with Dumping Syndrome. Eu J of Gastroenterology & Hepatology, 1989. 1:125-8.

14.	Freedman MR. 2000 Spectral analysis of Epigastrographic Signals for the assessment of gastric motility in Humans, MSc Thesis, University of Surrey.

15.	Li Z-Y, Ren C-S, Zhao S, Sha H & Deng J. Gastric motility functional study based on electrical bio-impedance measurements and simultaneous electrogastrography. Journal of Zhejiang University-Science B (Biomedicine & Biotechnology) 2011.12:(12) 983-9