Computers in biology and medicine
-
This work proposes the application of neural network multi-models to the prediction of adverse acute hypotensive episodes (AHE) occurring in intensive care units (ICU). A generic methodology consisting of two phases is considered. In the first phase, a correlation analysis between the current blood pressure time signal and a collection of historical blood pressure templates is carried out. ⋯ A correct prediction of 10 out of 10 AHE for event 1 and of 37 out of 40 AHE for event 2 was achieved, corresponding to the best results of all entries in the two events of the challenge. The generalization capabilities of the strategy was confirmed by applying it to an extended dataset of blood pressure signals, also collected from the MIMIC-II database. A total of 2344 examples, selected from 311 blood pressure signals were tested, enabling to obtain a global sensitivity of 82.8% and a global specificity of 78.4%.
-
Doctors applying mechanical ventilation need to find the best balance between benefit and risk for the patient. Mathematical models simulating patient's reactions to alterations in the ventilation regime may be employed. ⋯ The interaction of model systems reveals qualitatively varying results depending on the complexity of the involved models. Realistic overlaying of respiratory and cardiovascular rhythms can be detected in blood gas concentrations.
-
Noninvasive ventilation is a clinical procedure that enables patients with chronic respiratory failure to reduce the work of breathing and to improve blood oxygenation. In order to attain such goals, the ventilation support is expected to be phase synchronized with the patient spontaneous breathing. Unfortunately, asynchrony events are not rare. ⋯ Hence the estimation of input-output and autonomous models from pressure and airflow time series is discussed and illustrated. Issues concerning the nonlinearity of the interactions and modeling assumptions are dealt with. The results presented include models obtained from airflow and pressure measurements of a set of patients.
-
In this paper, we present a novel active contour (AC) model for medical image segmentation that is based on a convex combination of two energy functionals to both minimize the inhomogeneity within an object and maximize the distance between the object and the background. This combination is necessary because objects in medical images, e.g., bones, are usually highly inhomogeneous while distinct organs should generate distinct image configurations. Compared with the conventional Chan-Vese AC, the proposed model yields similar performance in a set of CT images but performs better in an MRI data set, which is generally in lower contrast.
-
It is known that, in cochlear implants (CIs), the addition of noise to CI stimuli may enhance the stochastic firing in the auditory nerve (AN). The aim of this study was to investigate, by using a model of a fiber of the AN the influence of CI stimulation settings (i.e., stimulation rate and stimulus magnitude) and fiber spontaneous activity on the enhancement of stochastic firing when CI stimuli are combined with noise. Results showed that the stimulation rate had an effect on the enhancement of stochastic firing in the AN, whereas the stimulus magnitude and the spontaneous firing rate had no influence.