[excerpt]
I believe that one of my most important contributions to neuro-gastroenterology was the development of new ways of graphically representing the mechanical events of intestinal motility. This has made motor patterns accessible via better visualisation and quantification. For me, the idea of regarding motor events as states of dynamic stability, portrayed in a graphical form, started during a trip to Shizuoka in 1993. I attended the International Conference on Gut Hormone, as a guest of Noburu and Shizuko Yanaihara. Walking in the city of Shizuoka, I saw a fountain made of a rectangular block of dark stone with horizontal indentations.
Water flowed from the top, sliding down on the wall face, appearing to form transient, but stable shapes which appeared to migrate down the stone to the collecting tray before being circulated back to the top by a hidden pump. I was fascinated by this sequence of ever- changing transient shapes produced by the simple flow of water over an irregular surface. It reminded me of the temporary stability of transiently living organisms. The figures originated with the flow of water (life) interacting with a surface, to form shapes which remain stable for a while, then vanish at the bottom (death).
This visual metaphor also triggered an idea that the transient events resembled the dynamic motor patterns of the intestine. I could see a relationship between the morphology of functional states which changes over time. The resulting shapes represent both anatomy and physiology at the same time. Still images of the watery shapes correspond to anatomical snapshots. Recording the changes of the patterned light over a vertical line would generate a trace of changes over time at a single location, similar to physiological recordings. A moving film of the events would contain information about both space and time. Furthermore, a series of time traces, laid side by side, would reconstruct both anatomical and physiological features simultaneously. The implication was clear; the difference between anatomical images and physiological traces is simply down to which part of the record they sample. The full record itself would include both spatial and temporal dimensions in a 4-dimensional array. Events with apparent beginnings and ends, like the water shapes of the “Fountains of Shizuoka”, would be better thought of as 4-dimensional objects.
Similarly, events such as peristalsis, which begin and end, are four-dimensional. This idea took off after Grant Hennig, a young PhD student, joined me. My ideas from the fountains of Shizuoka merged with his computer skills. We developed a way to extract the diameter at a single point along the intestine while it was undergoing peristalsis. Many such lines could be plotted with time in one dimension, with the spatial dimension, representing oral to anal (Hennig et al 1999). We produced a spatio-temporal map of the changes in diameter during peristalsis. A key development was the decision to transform diameters into values of gray and plot them along the segment over time.
This created spatio-temporal maps with transient objects that appeared to moved in time and space. This way of graphically portraying physiological events revealed a new way to represent intestinal motor patterns. A section of such a map, at a particular time, generates a spatial image (an ‘anatomical silhouette’) while, over time, a vertical line of the changes in diameter at a particular point generates a physiological trace. Being a digital map, measuring the propagation of contractions and the frequencies of events became easy.
…
The studies of the neural architecture of the enteric neural circuits were approaching completion and the functional classification of enteric neurons and their connections was making rapid progress. I thought that it was time to put our findings to the test. Modelling could be used to test whether defined anatomical and functional physical features could be integrated to generate digital simulations which mimic the real events. We took the approach to modelling developed by Sten Grillner in Sweden to simulate the swimming of lampreys (Ekeberg et al 1999). We published preliminary studies on intestinal peristalsis (Randawa et al 1995; Randhawa et al 1996).
However, it soon became clear that modelling the mechanical events of the intestine from the properties of individual cells was doomed to fail. Denis Noble had been attempting to model the heart, where much more detailed data was available. The initial approach of Hodgkin & Huxley in their Nobel prize-winning work had been successful in modelling the biophysical properties of single nerve cells. Noble faced the problem of computing capacity in modelling biological systems, particularly in going from cellular models to full organ models (Noble 2007). Some of our colleagues more versed in mathematical modelling managed to simulate some of the enteric neural circuits (Thomas et al 2004), although these models were not widely developed further.
Noble recently concluded that analytical models could not resolve the highly non-linear nature of the cellular mechanisms underlying organ behavior and that multiple levels of causations had to be included in the conceptual process (Noble 2012).
The dream of modelling complex interactions between enteric circuits, activated by mechanical stimuli and resulting mechanical events, was doomed to fail if concentrated on increasing the details of cellular properties. I turned my attention towards simplifying neural and mechanical events enough to provide a more conceptual model of intestinal motility.