A descriptive method for automatically analysing uterine forces during labor.

Abstract

In recent years many authors have considered evaluation of uterine contractility by means of analysis of amniotic fluid pressure variations. The reason for this is the introduction in obstetrical management of equipment for the recording of amniotic fluid pressure variations which is a powerful facility for clinical investigation, particularly, if it is considered that previous research has been based upon clinical evaluations and measurements carried out by extra and intra-uterine manometers. Important results of these studies are the definitions of two labor indices, the Montevideo unit (defined as the sum of the maxima of the pressure signal of all uterine contractions during a time interval of ten minutes) and the Alexandria unit (which also takes into account the duration of the contraction:this unit corresponds to the Montevideo unit multiplied by the mean value of the duration of the considered contractions). It is very difficult, however, in fact almost impossible, to summarize the evaluation of labor as the only indes. Many parameters which characterize the shape, frequency and tone of contraction patterns, may be themselves important and could indicated the presence of pathological elements. At present it is not well known which parameters are useful. Since many different sets of parameters can be chosen to characterize the signal, the first point is to determine which is more important clinically. The aim of this paper is to study a procedure of intrauterine pressure signal analysis, which allows the description of uterine contraction by a set of parameters that is interesting for research purposes during labor and suitable for continuous monitoring. The set of parameters was chosen in order to repeat quantitatively the qualitative analysis which a physician can make by a simple inspection of the pressure signal oscillogram. Visual examination of the tocograph results is a clumsy means for evaluating labor performed by an experienced physician. It may appear that the pressure starts "suddenly" (or not) to increase and rises "quickly" (or "slowly"), then it maintains "high" (or "low") mean values for a "long" (or "short") time, with "smooth" (or "irregular") behaviour, then it decreases "quickly" (or slowly, until "suddenly" (or "smoothly") it reaches its basal level, whd the beginning of another uterine contraction is "long" (or "short", during this time the pressure is "almost" constant (or has a "drift" and/or presents "many irregularities"). Therefore the 25 chosen parameters replace quantitatively the above qualitative evaluations. These parameters are not put forward to indicate directly the presence of pathological facttor, but summarize all the interesting clinical aspects of the contraction waveform for research and monitoring purposes. The parameters are extracted using a pattern recognition algorithm based on a structural representation of the contraction (Tables I, II, V)...