[Zurück]


Habilitationsschriften:

E. Kaniusas:
"Multiparametric Physiological Sensors";
Technische Universität Wien / Fakultät für Elektrotechnik und Informationstechnik, 2006.



Kurzfassung englisch:
Today the physiological sensors follow the rule that a single physiological parameter is attained per single sensor. In many cases, a single parameter may not be sufficient to determine a physiological abnormality. Thus two or more sensors might be needed to establish a reliable diagnosis. For instance, the diagnosis of sleep apnea, i.e., abnormal cessation of respiration during sleep, implies monitoring of a complete set of sleep and respiratory variables by at least eight different sensors distributed over the whole body. The large number of necessary sensors makes the observation expensive, error-prone, and disturbs both patient and even physiological parameters of interest.
The present work introduces multiparametric sensors which, in contrast to the common sensors, allow of the synchronous monitoring of a multitude of physiological parameters per single sensor. The attained multiparametric monitoring is less obtrusive and effectively addresses the economic imperatives, which is especially relevant for the long-term observation of patient physiologic state. However, it can not be expected that the multiparametric sensors will replace the established sensors anytime soon or even the stethoscope bearing clinician, but that may significantly expand the diagnostic capabilities in medicine and may have a broader impact on society, such as improved access to sleep apnea diagnosis.
The work shows that the multiparametric sensors can be established by the use of novel sensor concepts, optimised sensor locations as well as advanced signal processing methods, a priori knowledge about signal characteristics, and their mutual interrelations. In particular, the multiparametric approach is demonstrated by the use of three different sensors which operate on acoustical, optical, and mechanical basis:
. body sounds sensor,
. tissue absorption sensor, and
. skin curvature sensor.
The body sounds sensor for the registration of heart, lung, and snoring sounds mutates to a true multiparametric device if applied on an optimised chest region in combination with advanced sound processing methods and a priori knowledge about sound characteristics. This sensor allows the synchronous assessment of a numerous physiological parameters and events: cardiac activity, respiratory activity, breathing obstruction, apneas, and apnea types.
The tissue absorption sensor which measures light absorption of tissue at different wavelengths sets the second example of a multiparametric sensor if applied on a novel chest region. Theoretical investigations concerning the interaction of light propagation and physiological activities serve as a basis for a specific signal processing. As output of the processing, three parameters are offered, namely, cardiac activity, respiratory activity, and the changes of the arterial blood oxyhemoglobin saturation.
Finally, the multiparametric monitoring is demonstrated with a novel skin curvature sensor. It registers local deformations of the skin, which reflect a variety of physiological activities, the composition of the activities strongly depending on the skin region. Thus the sensor can be applied in a very versatile way, from the eyes down to the calf. For instance, the registration of the skin deformation on the neck over the carotid artery reflects simultaneously cardiac activity, respiratory activity, and even blood pressure changes. The focal point here is the novel approach in the assessment of the blood pressure characteristics causing minimum disturbance to the patient.
Since the physiological sensing and monitoring represent an interface between the applied technical equipment and human patient, an overview is given about the technical and application aspects regarding each type of the multiparametric sensor. In addition to the literature review, the overviews are substantially supplemented by own research data. In particular, multiparametric issues are considered in detail, including mutual dependences of physiological activities and cross-sensitivities of estimated parameters.

Erstellt aus der Publikationsdatenbank der Technischen Universität Wien.