New Aspects and Dimensions in Fetal Monitoring

Click here for Audio/Video Version  *requires RealPlayer - free download

Hans van der Slikke, MD: "It's September of 2001, and we're in Barcelona at the 5th World Congress on Perinatal Medicine. Next to me is Professor Domenico Arduini - welcome."

Domenico Arduini, MD: "Thank you. Welcome to you here in Europe in Barcelona."

Hans van der Slikke, MD: "Thank you. Professor Arduini is the Head of the Department of Gynecology and Obstetrics at the University in Roma, and we will talk about new developments in fetal monitoring. You had a presentation where you talked about your new system - the Fuzzy Logic, and this could add a lot of new items to fetal monitoring and a lot of new dimensions to it. Could you please explain this difficult topic as clear as possible?"

Domenico Arduini, MD: "Yes, the problem is that in the western countries we continuously have a problem with the cost of the assistance of the patients so the aim of our research is to try to find something very easy and very low cost. So in order to do that we use cardiotocography instead of the ultrasound scan because cardiotocography is something that's very low cost and you don't need personnel that's very well trained. Because of that we changed the hardware of one cardiotocographer with the help of Agilent (before Hewlett Packard) and now we will recognized the possibilities to study the heart rate and we now receive a signal every 350 milliseconds or 500 milliseconds. We have the RR-interval absolutely pure and we can use with the RR-interval similar to ECG with all new mathematical approaches like approximate antropy or like delayed function, decay function, or power special analysis. With that we found there's plenty of variables to be considered to study the fetal behavior and its well being. On the other hand, unfortunately we find one variable for the fetus. We have different variables for the different diseases so it's necessary to consider many variables for the different diseases of the fetus. Another problem is that each of these variables can change dramatically during the gestational age so it's possible that you have for IUGR fetuses one variable that can be useable in early gestation but is absolutely not useable in late gestation. Because of the multi-variables that come into play it is difficult to determine which patient's disease you will tackle and which gestational week has the best variables that you should consider."

Hans van der Slikke, MD: "So this makes it almost impossible for you to know which variable you should use."

Domenico Arduini, MD: "Yes, absolutely, it was our biggest problem because we have to consider all these incredible variables so for that we moved to an artificial neural network. Later on we used the system that's called Fuzzy Logic, it's used in different fields but not in medicine since eight or ten years ago. In fact, the first application of Fuzzy Logic was in the pediatric area and Fuzzy Logic gives the possibilities to an artificial network to decide on how many variables are to be used for a specific fetus. It determines the possibilities for the system to recognize without knowing beforehand if the fetus is at risk to be IUGR, at risk to have some problem if the mother is diabetic, or is at risk if the mother has a hypertensive disorder."

Hans van der Slikke, MD: "I see. Do I understand you properly that in this way, instead of seeing with the ultrasound if this is a growth retarded fetus you can discover this with your CTG?"

Domenico Arduini, MD: "This is a goal of our research, actually, in a small population we tried this in 450 patients and about 50% were high-risk patients. In a small population, yes, we have a good sensibility of the method only for these three diseases - for diabetes, hypertension, and IUGR depending or not depending from hypertension. We hope that in the future increasing the number of cases and increasing the capability of the system to understand (because as you know artificial intelligence needs to have its own experience) and ameliorate its own results if it continues to grow with its own experience. Now we've arrived at about 1,300 cases and we continue to store all these cases in it's own intelligence. We are trying to see if the results can ameliorate but actually the results are very impressive."

Hans van der Slikke, MD: "I understand this is still in the phase of research and not, let's say, yet for practical use in a general hospital."

Domenico Arduini, MD: "No, absolutely not, we are still in research because as I told you before, in the beginning it is necessary to change the hardware of the cardiotocography machine. It's very easy, it is not complicated but it is a necessity that the industry can understand and this is a first step. Secondly, we have to be sure that the false-negative is not so high because if it isn't this could be a very dangerous system. We need to reach about 10,000 or 15,000 cases before we think we can use it for clinical purposes. But now we are preparing an European multicenter trial for that and we hope that in two or three years we can reach this number."

Hans van der Slikke, MD: "We're looking forward to the results of this new trial. Thank you very much, Professor Arduini, for this interview."

Domenico Arduini, MD: "Thank you very much."