Fully automated analysis of colonic manometry recordings

Abstract

The purpose of this study was to develop a computer program for fully automated analysis of all presently known motor patterns in human colonic motility recordings. Colonic pressure recordings obtained from 24 healthy volunteers were used. Algorithms were developed for the detection and numerical analysis of five types of pressure waves: antegrade, retrograde, simultaneous, high-amplitude and isolated pressure waves. Furthermore, periodical motor activity was quantified. Validation was performed by comparison with visual analysis by two experienced observers. Patterns recorded during day- and night-time were compared using multiple-factor analysis of variance with Bonferroni correction. Automated analysis correlated well with visual peak detection (r = 0.98, P <0.01) and detection of antegrade pressure waves (r = 0.98, P <0.01). Most motor patterns showed a diurnal variation. During the night, prevalences of antegrade (938 vs 455; P <0.05), retrograde (112 vs 81; P <0.05), high-amplitude (12.9 vs 1.3; P <0.05), isolated pressure waves (1114 vs 765; P <0.05), and periodic motor activity were decreased (7.33 vs 4.47%; P <0.05). However, when expressed as percentage of absolute numbers of pressure waves, prevalences remained constant. In conclusion, fully automated analysis of all hitherto described colonic motility patterns is feasible. During the night, overall wave prevalences markedly decreased, but the distribution over the various motor patterns was preserved.

Abnormal Colonic Propagated Activity in Patients with Slow Transit Constipation and Constipation-Predominant Irritable Bowel Syndrome

“…Eighteen age-matched healthy volunteers (9 males, 9 females, 36–50 years old) were recruited as a control group; the controls were not matched by gender, since propulsive activity is not influenced by this variable [22]. No volunteer complained of abdominal pain, abdominal distention, or disturbances in bowel habits.…”

“…The HAPC were defined as pressure waves which (1) were observed at three ports sequentially with a delay of 1–10 s between peaks seen at adjacent sites, and (2) had amplitudes of more than 50 mm Hg absolute pressure at all three sites [22, 28](fig. 1).…”

Classification of normal and abnormal colonic motility based on cross‐correlations of pancolonic manometry data

“…Two notable approaches have been attempted. The first of these developed software that attempted to match human visual analysis, 19 and although this development was successful, all it really achieved was the detection of events that had already been defined manually, in a faster and arguably more objective fashion. Such an approach, however, is unlikely to define patient‐specific motility patterns.…”

“…Automated analysis of colonic motility data could potentially aid our ability to differentiate aberrant colonic contractility in patients from that of healthy controls. However, if the automated approach is based upon pattern‐recognition algorithms, which in turn are based on our own preconceived ideas of what a propagating event is, then the output will simply match our visual attempts 19 . Another approach published recently, utilized independent component analysis, which attempts to separate noise from genuine motility patterns, and then classify categories of rhythmic events 20 .…”

Evaluation of artificial neural networks in the classification of primary oesophageal dysmotility

“…In further studies more advanced approaches to signal processing and computer programming could be applied; for example automated pre-processing packages [26,27], cubic B-splines [28] or largest Lyapunov exponents [10]. In addition, ''unsupervised'' ANNs, i.e.…”