In this paper we present several statistic gradient algorithms from literature to solve the Principal Component Analysis (PCA) problem. We used a linear artificial neural network forming the basis of the implemented algorithms which is a neat way for on-line computation of the PCA expansion. As convergence is a key-aspect of these algorithms and is cru-cial for the usefulness in particular applications, we compared the different learning rules with respect to their suitability in ECG signal processing. Recent studies have shown, that a surrogate respiratory signal can be derived from single-lead ECGs by applying PCA. Since the traditionally applied closed-form computations of PCA are numerically demanding, it seems promising to resort to an adaptive approach when dealing with changing environments like the ECG.

In this paper a computationally efficient algorithm for continuous blood pressure curve segmentation is presented. It uses only methods in the time domain and can distinguish between systolic, diastolic values and values of calibration steps caused by the continuous blood pressure measuring technique or values of other artefacts. The detection of local extremes, necessary for systolic and diastolic points, is performed with smoothed first and second derivations. An adaptive threshold approach sorts out most of the false extremes, not belonging to the valid blood pressure curve. But only the following plateau detection is suitably reliable to detect local extremes which lie within a calibration step.

Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Assigning the hemocompatibility of small vascular implants is one of the great challenges in biomedical engineering. Due to the fact, that there are no widely approved test setups 1 , we decided to developed a modified CHANDLER-Loop system for dynamic in vitro hemocompatibility tests. The setup allows simultaneous testing of about 30 tube rings with an inner diameter of 2,4 mm. Cardiovascular implants can be placed into these tube rings. After filling them with anticoagulated blood and closing them to loops, they were placed on a rotating disc. While the loop is rotating, the blood remains in the lower part of the loop. The relative movement between the foreign surface and the blood sample induces a blood flow without the need of mechanical pumping. A first test series with the new test setup had been conducted to achieve a highly hemocompatible tube material. The results were compared to hemocompatibility tests done by LEMM, mentioned in the ISO 10993-4 2,3 . We could prove, that Tygon S50HL ® and Polyurethane achieve similar hemocompatibility values. In addition, we did in vitro tests of bare metal stents. We recorded a difference in the index of thrombocytes of about 17 % between the loops with three stents and the control. During further experiments the difference decreases to 8 %, which might be due to adhesion processes. The ability to measure these processes demonstrates the high sensitivity and very low background activation of the test setup. This can be assured by the recorded index of hemolysis, which had been less than 0,8 % in all our experiments. By developing and improving our modified CHANDLER-Loop system, we were able to establish a dynamic in vitro test setup for the hemocompatibility testing of small vascular implants. Because of its extremely low background activation and a high variability, the system serves as a benchmark for upcoming test setups.