Open Access
Add to BibSonomyAbstract
Fiber optic catheters for the diagnosis of gastrointestinal motility disorders are demonstrated in-vitro and in-vivo. Single element catheters have been verified against existing solid state catheters and a multi-element catheter has been demonstrated for localized and full esophageal monitoring. The multi-element catheter consists of a series of closely spaced pressure sensors that pick up the peristaltic wave traveling along the gastrointestinal (GI) tract. The sensors are spaced on a 10 mm pitch allowing a full interpolated image of intraluminal pressure to be generated. Details are given of in-vivo trials of a 32-element catheter in the human oesophagus and the suitability of similar catheters for clinical evaluation in other regions of the human digestive tract is discussed. The fiber optic catheter is significantly smaller and more flexible than similar commercially available devices making intubation easier and improving patient tolerance during diagnostic procedures.
©2009 Optical Society of America
1. Introduction
Gastrointestinal disorders have a significant impact upon health care resources. In the USA during 2002 an estimated 31 million clinic visits were made by patients seeking treatment for conditions such as abdominal pain, dysphagia (the inability to swallow), irritable bowel syndrome, diarrhea, constipation, nausea and vomiting [1]. While the majority of patients will be able to attain suitable treatment for their medical condition there are a number of gastrointestinal disorders for which treatment remains unsatisfactory. For example while constipation is often perceived as a benign, easily treated condition it has significant adverse impact on health-related quality of life [2] and at least 36% of those presenting to the clinic subsequently fail non-surgical therapies (diet, bulking agents, laxatives, biofeedback) [3]. Such patients account for 13.7 million days of restricted activity and 3.4 million days of bed disability in the US [4], costing the US health care system an estimated $US235M per year [5]. While these figures concentrate on constipation the monetary values attributed to the direct and indirect costs of irritable bowel syndrome reach as high as $US30billion [6].
Improved therapies for the treatment of gastrointestinal disorders can only stem from a clearer understanding of the phenomena underlying the condition. Our current knowledge of in vivo motor patterns in the gastrointestinal tract is largely derived from manometric studies: that is the recording of rhythmic contraction of smooth muscles that propel contents through the digestive tract (peristalsis). For example in patients with severe constipation, available evidence indicates abnormal colonic contractility as a probably cause [7, 8]. It is further believed that in health, the motor patterns have a short range coherence spanning 3 – 10 cm along the gut that contributes to the effective propulsive mechanism [9]. However, our ability to detect these motor patterns is largely dependant upon the spatial resolution of recording sites on the manometric catheter that is being used. For example, in most colonic catheters the spacing of pressure sensors along the device is >7.5cm, so any short range coherence is missed and our ability to accurately define normality and subsequently abnormalities is severely impeded.
To combat this problem high-resolution manometry (HRM) using catheters with 10 mm spacing between recording sites can be adopted. This technology is now regularly performed in the pharynx, oesophagus, stomach, duodenum, jejunum and anorectum [10–13]. HRM recordings offer a far greater understanding of short range motor patterns because the inter-sensor spacing is now less than the typical range of the motor patterns being observed. These recordings help to highlight important physiological and pathophysiological motor patterns that are not as evident with standard low-resolution manometry.
There are two distinct technologies currently being employed for HRM recordings; solid-state pressure sensors [14, 15] and water-perfusion sensors [16, 17]. Although both can provide distributed force/pressure recordings adequate for diagnosis of motility disorders, they both suffer from the need for increasing complexity as the sensor count rises and, of necessity, an increase in catheter diameter as more sensors are added. Each has its own advantages and limitations [18]. Water-perfused catheters are robust, inexpensive, relatively flexible and amenable to heat sterilization by autoclave. However, they require external bulky water perfusion pumps, manifolds and bulky external transducers. Perfusion manometry also suffers from added compliance resulting in a lower pressure rise rate (i.e. lower frequency response) than do solid state sensors [19]. Solid state devices have the advantage of simplicity and very high frequency response (only relevant in the pharynx, and not particularly important in the remainder of the gut. Solid-state devices tend to become less flexible as the number of sensors, and hence the number of electrical conductors, increases making intubation harder to tolerate. Solid state sensors are not autoclavable, are costly and fragile; all of which adds to the health care cost burden relating to their use.
Due to the increasing complexity and catheter diameter, current practical limitation of the number of sensor sites are ~36 for the solid state technology and ~20 for the water perfused technology. So if these recording sites are spaced at 10 mm intervals the total length of the recording device is <35cm, and while such distances are adequate for recording detailed data from regions such as the oesophagus, they are insufficient for recording data from the colon which can span 1.2m. The limited flexibility of solid state devices also limited their use in the large and small bowel, while the discharge of water into the gut region under investigation from the water perfused devices may at times perturb the phenomena under investigation.
To overcome the current limitations of manometric devices we have adapted fibre optic strain gauges based on fibre Bragg gratings to assemble catheters with similar functionality to the existing solid state and water perfused devices [20, 21]. These catheters have a smaller diameters and significantly increased flexibility, in comparison to solid state catheters, making them easier to introduce into the human digestive tract. Furthermore, the application of Wavelength Division Multiplexing (WDM) techniques allows the simultaneous interrogation of a large number of sensors along a single fiber, thus making it possible to extend the total span of the HRM sensing region, without increasing the girth of the resulting catheter.
These catheters offer great promise for recording in the colon and other extended regions of the gut however since placement of catheters into the colon is time consuming our initial studies have concentrated on recordings in the human oesophagus.
In this work we present both in-vitro and in-vivo validation of a series of single element fiber optic catheters fabricated using this technology against an equivalent commercially available solid-state manometry catheter. Initial in-vivo results from a 32-element fibre optic HRM catheter are also given and show that the technology is capable of picking up diagnostically significant features of peristalsis.
2. Catheter design and data acquisition
The catheters were fabricated from either single or multiple fibre Bragg grating (fbg) elements written into a continuous length of single mode fiber [22]. Each fbg was rigidly fixed to a localized pressure sensitive structure consisting of a rigid metallic substrate and a flexible diaphragm. The fbg structures were written between 815 and 850 nm and their spectral widths were defined by the maximum gauge length of the substrate which was set at 3.6 mm by the design requirements of the catheter. The fbg element length was therefore set to 3 mm which resulted in their full widths (to first zeros) being approximately 0.6nm.
In use, changes in ambient pressure cause the diaphragms to flex sideways against the sections of fibre containing the respective fbg element, which in turn modified the reflected Bragg wavelength of that element. The local ambient pressures were then inferred by comparing the reflected wavelengths to values determined during initial calibration of the device. To allow adequate dynamic range of the pressure sensors the spectral peaks were separated by 1.3 nm, which in turn limited the number of sensors to approximately 32 using a commercially available ‘BlueBox’ interrogator unit from IPhT [23]. By forming the fbg array from spectrally separated elements and detecting in the spectral domain using the BlueBox, it was possible to determine the ambient pressure variations along the length of the catheter simultaneously by capturing the reflected spectra in real time (this approach is known as wavelength division multiplexing).
A schematic view of the fbg sensing region is shown in Fig. 1 and a simplified layout of the associated test and measurement hardware is shown in Fig. 2.
To form the catheter the sensor array was inserted into an outer sleeve of silicone rubber. This outer sleeve was sealed at the distal end with a room-temperature-vulcanizing (RTV) rubber plug and the proximal end was finished with an FC/APC optical connector to facilitate easy connection to the data acquisition system.
3. In-vitro calibration
Prior to intubation of the catheter into the patient, the catheters were calibrated against an electronic pressure transducer. The expected variations in pressure experienced in the human oesophagus are in the range 0 to ~200 mmHg so during calibration the pressure was varied over a slightly greater range to allow viable curve fits to the data to be made over the range of interest. These curve fits were then used to infer the ambient pressure felt at each sensor location during subsequent in-vitro and in-vivo data acquisition. Figure 3 shows a typical Pressure-Wavelength curve recorded during calibration, indicating that the response of the pressure sensors was close to linear. The standard deviation between the recorded data and the curve fits were <3.1 mmHg over the range 0 to 150mmHg.
Fig. 3. Typical calibration curve for one of the fbg pressure sensors. Calibration was carried out with both increasing and decreasing pressures. The black line is a linear fit to the data.
4. In-vivo measurements
4.1 Single-element validation
In-vivo measurements were performed in the oesophagus of healthy human controls by recording a series of 10 ml water swallows. Since biometric data from human peristalsis is highly variable from swallow to swallow and also from subject to subject it was necessary to validate the in-vivo catheter performance by direct comparison with a known solid state catheter (Gaeltec, Dunvegan, Scotland). To do this in a statistically significant manner, a batch of 10 single-element fiber optic catheters were fabricated and were in turn taped to the solid state catheter so that the pressure sensing regions were collocated as shown in Fig. 4.
After fasting for at least 4 hours, 10 healthy volunteers were intubated via the nose with both catheter assemblies. While positioned in an upright seated position, the volunteers were asked to perform three series of 10 controlled water swallows (each of 10 mls volume) with the sensing regions located in the proximal, mid, and distal regions of the esophageal body. This gave a total of 300 swallows that, after obvious outliers were removed from the data, were analyzed statistically using a Bland-Altman analysis [24]. Figure 5 shows the comparison of peak amplitudes recorded by the solid state and fibre optic devices that resulted in a concordance correlation coefficient of =0.992 and a 95% confidence interval of 0.989 to 0.994 indicating a high degree of correlation between the fiber-optic and solid-state recordings.
Figure 6 shows a single swallow recorded by the catheters and shows the close correlation between the recorded measurements. The high frequency component in the quiescent regions of the trace from the fiber-optic catheter is the heart-beat of the volunteer due to the close proximity of the descending aorta to the esophagus.
4.2 Multiple-element measurements
Following the successful validation of single-element fibre-optic devices against the solid-state catheter, a 32-element fibre-optic catheter was fabricated with the individual pressure sensors spaced at 10 mm intervals along the distal region of the catheter. The outer diameter of the catheter was slightly less than 3 mm and the sensor region was highly flexible allowing easy nasal intubation and improved tolerance of the device by the patient during the clinical evaluation. The sensing region of the catheter is shown in Fig. 7 indicating the high degree of flexibility of the sensor region.
Each sensor along the length of the catheter was calibrated in the manner described above and the catheter was intubated into a healthy volunteer. The outputs of each sensor were viewed as line plots and also the instantaneous outputs from each element were interpolated using a cubic-spline fitting routine to give a real time image of the pressure variations along the full length of the esophagus. These interpolated snap-shots were recorded on a scrolling intensity plot with sensor position along the Y-axis and evolving time along the X-axis. This allows the diagnosing clinician to view the recorded output from the catheter in a single 2D image known as a spatio-temporal plot (S-T) [10, 12, 15]. To demonstrate the improved nature of the information gained by HRM catheters in general and specifically from S-T plotting, Figure 8(A) shows line plots from a standard 8-element water-perfused manometry catheter with sensing regions spaced at 5 cm; and Figs. 8(B) & 8(C) show the nested line plots from each individual sensor of a 32-element fibre optic HRM catheter the S-T plot of a typical swallow respectively. The details of the S-T plot will be given more fully below.
Fig. 8. Output from (A) Line plot outputs from an 8-element water perfused manometry catheter demonstrating the reduced information available from low-resolution manometry devices (B) line plots from the 32-element fibre optic catheter during a typical swallow showing outputs from each individual sensor; and (C) an interpolated spatio-temporal plot of the same data.
The relevance of the 32-element, 10 mm pitch design for esophageal manometry was that this gives an overall length to the sensing region capable of spanning the full length of the esophagus from pharynx to stomach and including upper and lower sphincter regions. The upper and lower esophageal sphincters (UES and LES respectively) are the regions of high muscular contraction that control the transit of material into the esophagus and stomach respectively. These regions serve a “valve-like” function, undergoing abrupt relaxation and opening during critical phases of swallowing to permit passage of food boluses through this region into the oesophagus and stomach respectively. The elevated muscular contraction recorded in these regions allows a direct method of accurately placing the catheter during intubation such that the peristaltic contractions running the full length of the esophagus can be readily recorded by the sensor array. The catheter was passed via the nose until the measured high tension regions corresponding to the UES and LES were visible at the top and bottom of the S-T plot as bands of color running across the image, as seen in Fig. 8(C). In this figure the UES is clearly visible, whereas the LES is less clear but still adequate for accurate positioning of the catheter.
Once the catheter was correctly positioned in the esophagus, the patient was again asked to carry out a series of controlled 10 ml water swallows so that an overall image of peristalsis, and other clinically significant features could be recorded. This provided enough information to determine the efficacy of peristalsis.
4.3 Interpretation of HRM data
In Fig. 8(C) the location of the UES is clearly visible between elements 2 and 6 and LES between elements 28 and 30. The swallow is initiated at approximately 1.5 seconds which triggers the relaxation of the UES and LES. The peristaltic contraction is then indicated by the diagonal band of color running from top left to bottom right. As the wave of peristalsis passes each sphincter region the normal muscular tension returns sealing each end of the esophagus.
In addition to the overall image of the swallow, further diagnostic information can be gleaned from the individual traces recorded by the sensors. Figure 9 shows the time varying recording from sensor #3 located within the UES, and shows a number of distinct and clinically significant features.
Fig. 9. Time-varying output from sensor #3 located within the UES. Details of the plot are given in the text.
As the swallow is initiated at ~1.5 seconds, the natural tension in the sphincter relaxes in preparation for the transit of the bolus (point A). The sharp rise in tension seen at point B indicates the transit of the peristaltic wave, and finally after point C the normal tension of the sphincter returns to seal off the top of the esophagus.
Similarly, Fig. 10 shows the data from sensor #21 located just above the LES. In this instance there are two distinct regions. Firstly, as the liquid bolus reaches the constriction of the LES at approximately 4 seconds it pools in the lower esophagus and covers the sensor. This causes a small rise in pressure at point A due to the hydrostatic pressure of the liquid bolus. The peristaltic wave then impacts on the region of the sensor at point B forcing the bolus through the LES and into the stomach. The initial rise in hydrostatic pressure is known as the intra-bolus pressure.
Fig. 10. Time-varying output from sensor #21 located above the LES. Details of the plot are given in the text.
The full S-T plot is useful to determine the overall efficacy of peristalsis whereas the individual line plots can be used to determine the correct operation of discrete sections of the esophagus. The combined data provides a powerful means of interpreting both healthy and diseased states of gastrointestinal motility.
Conclusion
A series of fiber optic manometry catheters have been fabricated using fbg based pressure sensors. In a controlled in-vivo clinical trial, a batch of 10 single element fiber optic catheters were directly compared with a commercially available solid state manometry catheter and were shown to provide a substantially equivalent response. A 32-element catheter was then fabricated using an array of wavelength separated fbg’s monitored in real time using WDM interrogation techniques. The catheter was intubated in a healthy volunteer and used to record a number of controlled swallows. The results demonstrate the ability of the fiber optic catheter to record peristalsis running from pharynx to stomach and also individual features of the peristaltic contractions.
The catheters had an outer diameter of less than 3 mm and were very flexible, allowing easy nasal intubation and relative comfort for the patient during subsequent clinical diagnosis.
References and links
1. N. J. Shaheen, R. A. Hansen, D. R. Morgan, L. M. Gangarosa, Y. Ringel, M. T. Thiny, M. W. Russo, and R. S. Sandler, “The burden of gastrointestinal and liver diseases, 2006,” Am. J. Gastroenterol . 101, 2128–2138 (2006). [CrossRef] [PubMed]
2. C. Dennison, M. Prasad, A. Lloyd, S. K. Bhattacharyya, R. Dhawan, and K. Coyne, “The health-related quality of life and economic burden of constipation,” Pharmacoeconomics 25, 461–476 (2005). [CrossRef]
3. P. C. Rantis Jr., A. M. Vernava 3rd., G. L. Daniel, and W. E. Longo, “Chronic constipation--is the work-up worth the cost?,” Dis. Colon Rectum 40, 280–286 (1997). [CrossRef] [PubMed]
4. A. Sonnenberg and T. R. Koch, “Epidemiology of constipation in the United States,” Dis. Colon Rectum 32, 1–8 (1989). [CrossRef] [PubMed]
5. B. C. Martin, V. Barghout, and A. Cerulli, “Direct medical costs of constipation in the United States,” Manag. Care Interface 19, 43–49 (2006).
6. L. Chang, “Review article: epidemiology and quality of life in functional gastrointestinal disorders,” Aliment Pharmacol. Ther . 20 Suppl 7, 31–39 (2004). [CrossRef]
7. P. G. Dinning, P. A. Bampton, J. Andre, M. L. Kennedy, D. Z. Lubowski, D. W. King, and I. J. Cook, Abnormal predefecatory colonic motor patterns define constipation in obstructed defaecation,” Gastroenterol . 127, 49–56 (2004). [CrossRef]
8. S. S. Rao, P. Sadeghi, J. Beaty, and R. Kavlock, “Ambulatory 24-hour colonic manometry in slow-transit constipation,” Am. J. Gastroenterol . 99, 2405–2416 (2004). [CrossRef] [PubMed]
9. W. M. Sun, G. S. Hebbard, C. H. Malbert, K. L. Jones, S. Doran, M. Horowitz, and J. Dent, “Spatial patterns of fasting and fed antropyloric pressure waves in humans,” J. Physiol . 503, 455–462 (1997). [CrossRef] [PubMed]
10. R. B. Williams, A. Pal, J. G. Brasseur, and I. J. Cook, “Space-time pressure structure of pharyngo-esophageal segment during swallowing,” Am. J. Physiol. Gastrointest. Liver Physiol . 281, G1290–1300 (2001). [PubMed]
11. J. M. Andrews, D. G. O’Donovan, G. S. Hebbard, C. H. Malbert, S. M. Doran, and J. Dent, “Human duodenal phase III migrating motor complex activity is predominantly antegrade, as revealed by high-resolution manometry and colour pressure plots,” Neurogastroenterol. Mot . 14, 331–338 (2002). [CrossRef]
12. M. M. Szczesniak, N. Rommel, P. G. Dinning, S. E. Fuentealba, I. J. Cook, and T. I. Omari, “Optimal criteria for detecting bolus passage across the pharyngo-oesophageal segment during the normal swallow using intraluminal impedance recording,” Neurogastroenterol. Mot . 20, 440–447 (2008). [CrossRef]
13. M. P. Jones, J. Post, and M. D. Crowell, “High-resolution manometry in the evaluation of anorectal disorders: a simultaneous comparison with water-perfused manometry,” Am. J. Gastroenterol . 102, 850–855 (2007). [CrossRef] [PubMed]
14. J. A. Castell and D. O. Castell, “Modern solid state computerized manometry of the pharyngoesophageal segment,” Dysphagia 8, 270–275 (1993). [CrossRef] [PubMed]
15. J. E. Pandolfino, S. K. Ghosh, J. Rice, J. O. Clarke, M. A. Kwiatek, and P. J. Kahrilas, “Classifying esophageal motility by pressure topography characteristics: a study of 400 patients and 75 controls,” Am. J. Gastroenterol . 103, 27–37 (2008).
16. M. Fox, G. Hebbard, P. Janiak, J. G. Brasseur, S. Ghosh, M. Thumshirn, M. Fried, and W. Schwizer, “High-resolution manometry predicts the success of oesophageal bolus transport and identifies clinically important abnormalities not detected by conventional manometry,” Neurogastroenterol. Mot . 116, 533–542 (2004).
17. A. J. Bredenoord, B. L. Weusten, R. Timmer, and A. J. Smout, “Sleeve sensor versus high-resolution manometry for the detection of transient lower esophageal sphincter relaxations,” Am. J. Physiol. Gastrointest. Liver Physiol . 288 (2005). [CrossRef] [PubMed]
18. J. E. Pandolfino and P. J. Kaharilas, “American Gastroenterological Association medical position statement: Clinical use of esophageal manometry,” Gastroenterol . 128, 207–208 (2005). [CrossRef]
19. T. Omari, M. Bakewell, R. Fraser, G. Davidson, and J. Dent, “Intraluminal micromanometry: an evaluation of the dynamic performance of micro-extrusions and sleeve sensors,” Neurgastroenterol. Mot . 8, 241–245 (1996). [CrossRef]
20. T. Omari, J. W. Arkwright, S. N. Doe, N. Rommel, M. M. Szczesniak, P. G. Dinning, and I. J. Cook, “Manometry in the 21st Century: A Novel Fibre Optic Based Technology for Multi-Channel Intraluminal Manometry,” Gastroenterol . 130, A730 (2006).
21. J. W. Arkwright, S. N. Doe, M. C. Smith, N. G. Blenman, I. D. Underhill, S. A. Maunder, B. Glasscock, M. M. Szczesniak, P. G. Dinning, and I. J. Cook, “A Fibre Bragg Grating Manometry Catheter for In-vivo Diagnostics of Swallowing Disorders,” TuI-2, OECC/ACOFT 2008, Sydney, Australia , 7–10 July (2008).
22. C. Chojetzki, M. Rothardt, S. Ommer, S. Unger, K. Schuster, and H. R. Mueller, “High-reflectivity draw-tower fiber Bragg gratings-arrays and single gratings of type II,” Opt. Eng . 44, 060503 (2005). [CrossRef]
24. J. M. Bland and D. G. Altman, “Statistical methods for assessing agreement between two methods of clincal measurement,” Lancet 1, 307–310 (1986). [CrossRef] [PubMed]
