The work concerns the study of the possibility of using an artificial neural network to determine the gas pressure or liquid, in the flow system. The basis for determining the pressure is the view of the membrane, which is obtained discreetly from the vision sensor. The essence of the method operation consists of associating the fuzzy image of the marker placed on the membrane with the corresponding reference pressure value, which in the network learning process, is read from the standard pressure gauge. The test used a device allowing the measuring of gas pressure with an accuracy no lower than 2%. The operation of the artificial neural network is based on identifying the degree of blurring the marker on the examined views of the membranes and associating them with the pressure values. In the case when the membrane views cannot be uniquely qualified for the training set, the network acts as an interpolator and predicts the pressure value.
The paper investigates the influence of selection and operators of evolutionary strategy on the reconstruction of the shape of the pneumatic membrane surface of the extracorporeal cardiac support pump. Sets consisting of selection, mutation and crossing were assessed. The study was conducted in the context of optimizing the distribution of markers on the surface of the flaccid membrane. The arrangement of the markers is important from the point of view of modeling the shape of the membrane surface and ultimately determining the stroke volume. The experiments were carried out for a convex membrane with a known mathematical description. The value of the error of mapping the determined shape of the membrane with respect to the shape of the reference surface was assumed as the criterion of the assessment.
The work concerns the study of the possibility of using an artificial neural network to determine the ejection volume of pulsatile models of heart assist pumps. The research used new pump designs, significantly different from those used in terms of dimensions and the material from which the flaccid membrane was made. The basis for determining the ejection volume are the special features of the membrane view, which is obtained from the vision sensor. The essence of the method operation depends on associating the membrane view with the corresponding reference volume value, which during the network learning process, is read from the burette with an accuracy of ±0.5 ml. The operation of the artificial neural network consists in the identification of artifacts on the examined views of the membranes and associating them with the ejection volume values. In the case where the membrane view cannot be univocally qualified to the training set, the network acts as an interpolator and predicts the stroke volume value. Verifying the ability to determine the stroke volume by the neural network was performed in close-to-real conditions. In addition to the test results, the article presents new pump designs, the laboratory station and the course of the experiment.
In the paper the research results, which are a continuation of work on the use of image processing techniques to determine the membrane shape of an artificial ventricle, were presented. The studies focused on developing a technique for measuring the accuracy of the membrane shape mapping. It is important to ensure the required accuracy of determining the instantaneous stroke volume of a controlled pneumatic artificial ventricular. Experiments were carried out on the following type of membrane models: convex, flat and concave. The purpose of the research was to obtain a numerical indicator, which will be used to evaluate the options to improve mapping techniques of thee shape of the membrane.
In the article we presented results obtained during research, which are the continuation of work on the use of artificial neural networks to determine the relationship between the view of the membrane and the stroke volume of the blood chamber of the mechanical prosthetic heart. The purpose of the research was to increase the accuracy of determining the blood chamber volume. Therefore, the study was focused on the technique of the features that the image extraction gives. During research we used the wavelet transform. The achieved results were compared to the results obtained by other previous methods. Tests were conducted on the same mechanical prosthetic heart model used in previous experiments.
The paper presents the use of an artificial neural network in sensors application. The task is to determine the volume of the chamber. The tests were performed on a model of a chamber in a mechanical prosthetic heart. In the considered task the surface of the diaphragm is observed by a near-infrared band camera. The artificial neural network was used to determine the relationship between the real views of the diaphragm and stroke volume. The artificial neural network learning process and research results are presented in the article.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.