Investigation Into The Relationship Between Intestinal Slow Waves, Spike Bursts, and Motility

Abstract

The small intestine musculature consists of longitudinal and circular muscle layers, whose coordinated contractions and relaxations facilitate the breakdown and digestion of food via motility. Intestinal motility is governed by a multitude of regulatory mechanisms including neuronal, hormonal, myogenic means, and is associated with two bioelectrical events: (i) slow waves generated and propagated by interstitial cells of Cajal, and (ii) spike bursts which are believed to be calcium currents in the smooth muscle cells. The relationship between slow waves, spike bursts, and motility is of critical interest to the intestinal function, but is not clearly defined at the organ level. Recent high-resolution electrical mapping has significantly improved the experimental and clinical understanding of slow wave activity in normal and dysrhythmic states. However, a similar level of understanding has not been translated to spike bursts. Furthermore, electrophysiological studies are generally performed separately from motility measurements, and there is a need to integrate these two modalities to understand the electrophysiological basis of motility. A limited number of simultaneous electrophysiological and motility studies exist, but are typically performed in-vitro, and have utilised 1-dimensional electrode arrays and motility measurements. This thesis aims to investigate the spatiotemporal relationships between slow waves, spike bursts, and motility, via simultaneous high-resolution mapping of bioelectrical activity and motility in the in-vivo intestine. First, an experimental setup was developed to simultaneously measure the bioelectrical and contractile activity of exposed intestinal segments from anaesthetised pigs and rabbits in spatiotemporal detail. The bioelectrical activity was recorded using high-resolution flexible electrode arrays (16×8 configuration, 4 mm inter-electrode spacing). The contractile activity was simultaneously recorded using a cross-polarised camera setup (5-megapixel camera BFS-U3-50S5M-C, Blackfly S, FLIR, USA; fitted with lens HF25SA-1, Fujifilm, Japan; 70 × 59 mm field of view) at 20 frames-per-second, and was synchronised with the bioelectrical recordings. Slow waves were detected, and frequency, amplitude, velocity were quantified using existing methods. A new automated framework was developed for spike burst analysis. The framework included detection of spike bursts and clustering them into patches, validated by manual review. The spike burst patches were segmented into longitudinal, circumferential, and propagating circumferential patches based on the activation patterns of spike bursts. Spike bursts and spike burst patches were quantified by calculating frequency, amplitude, duration, patch size, energy, and velocity. A method based on free-form deformation was developed to map and quantify in-vivo motility patterns in high spatiotemporal detail. The deforming geometry of the intestine in video sequences was modelled by a biquadratic B-spline mesh. Then, Green-Lagrange strain fields were computed based on the change in geometry to quantify the surface deformations from motility. A nonlinear optimisation scheme was applied to mitigate the accumulation of tracking error associated with image registration. The strain error was maintained under 1 % and the optimisation scheme was able to reduce the rate of strain error by 97 % during synthetic tests. The algorithm was able to generate 2D strain fields to quantify the contractile activity across the surface of the in-vivo intestine including anisotropic contractions, and were able to simultaneously map the coordinated activity of the circular and longitudinal muscle layers during motility patterns. The level of contraction (strain), rate of contraction (strain-rate), frequency, and velocity of propagation were calculated to quantify motility patterns. The experimental setup and analytical methods were applied to investigate the spatiotemporal dynamics between slow waves, spike bursts, and motility in the in-vivo jejunum of 6 anaesthetised pigs and 3 rabbits. Two types of spike bursts were observed: (i) smaller morphology slow wave associated spike bursts that activated periodically with slow waves (0.1 ± 0.1 mV, 0.8 ± 0.3 s, 10.8 ± 4.0 cpm in pigs; 0.1 ± 0.1 mV, 0.4 ± 0.2 s, 10.2 ± 3.2 cpm in rabbits), and (ii) larger aperiodic independent spike bursts that were not associated with slow waves were observed in pigs (1.4 ± 0.8 mV, 1.8 ± 1.4 s, 3.2 ± 1.8 cpm). Spike bursts activated as longitudinal or circumferential patches with associated contractions in the respective directions. The level of contraction correlated with the amplitude, size, and energy of spike burst patches. The rate of contraction correlated with the amplitude, duration, size, and energy of spike burst patches. Segmental contractions of 16±9 % in pigs were spatially correlated with circumferential patches of independent spike bursts. Pendular longitudinal contractions of 19 ± 6 % in pigs, 12 ± 4 % in rabbits were spatially correlated with longitudinal patches of slow waves associated spike bursts. Propagating circumferential patches of independent spike bursts led to spontaneous peristaltic contractions of 36 ± 4 % in pigs, which propagated slower than slow waves (3.7 ± 0.5 mm/s vs 10.1 ± 4.7 mm/s slow-wave velocity, p = 0.007). Propagating circumferential patches of slow waves associated spike bursts led to cyclic peristaltic contractions of 17 ± 2 % in rabbits, which occurred at a similar frequency to slow waves (11.0 ± 0.6 cpm vs 10.8 ± 0.6 cpm slow wave frequency, p = 0.97), and propagated at a similar velocity to slow waves (14.2 ± 2.3 mm/s vs 11.5 ± 4.6 mm/s slow wave velocity, p = 0.162). The results demonstrated that spike burst propagation patterns dictated the resultant contractions. Spike bursts occurred coupled to slow wave activations, but also operated independent of slow wave activations, demonstrating that in the jejunum, slow waves were not always correlated to the contractile response. The experimental setup and analysis methods were also applied to investigate the electrophysiological and contractile changes during mesenteric ischaemia, in spatiotemporal detail. Experiments were performed on 5 anaesthetised pigs with the bioelectrical and video mapping techniques as described previously. First, the baseline activity was recorded, then the mesenteric vessels supplying to the intestinal segment were clamped to induce ischaemia (for 18.2±9.0 minutes), and again unclamped to record the activity during reperfusion (for 3.5 ± 1.0 minutes). Slow wave entrainment within the ischaemic region diminished, resulting in sporadic slow wave activations and a reduction in frequency from 12.4 ± 3.0 cpm to 2.5 ± 2.7 cpm (p = 0.0006). The deterioration of slow waves blocked the slow wave propagation across the ischaemic intestinal segment, and decoupled the distal slow wave activity from the proximal slow waves. During reperfusion, slow waves regained the normal rhythmic nature, increased to 11.5 ± 2.9 cpm, and propagated throughout the previously ischaemic segment. Spike burst frequency increased during ischaemia from 1.1 ± 1.4 cpm to 8.7 ± 3.3 cpm (p = 0.0003), activated as propagating and non-propagating circumferential patches, and caused a spasm of circumferential contractions. During reperfusion, the frequency of spike bursts decreased to 2.7 ± 1.4 cpm, and contractions subsided. The intestine also underwent tonal contraction during ischaemia, with the diameter decreasing from 29.3 ± 2.6 mm to 21.2 ± 6.2 mm (p = 0.0020). During reperfusion, the intestinal diameter increased to 27.3 ± 3.9 mm. The intestinal slow wave, spike burst, and diameter measurements were not statistically different for baseline and reperfusion (p > 0.05). The decrease in slow waves, increase in spike bursts, and the tonal contraction can objectively identify ischaemic segments in the intestine and could also be used to verify successful revascularisation during surgery. The work presented in this thesis improves the understanding of the relationship between bioelectrical slow waves, spike bursts, and motility in the intestine. In conclusion, spike burst propagation primarily dictates the motility patterns, and slow waves play a key role in initiating cyclic motility patterns by coordinating spike burst activations. This thesis also demonstrates that abnormal slow waves, spike burst activity, and abnormal contractile patterns could be used to diagnose gastrointestinal conditions such as mesenteric ischaemia. In addition, the methods developed in this thesis can be translated into other areas of the gut to investigate motility patterns and electrophysiological control.