Johannes Kropf

Validation of a Brachial Cuff-Based Method for Estimating Central Systolic Blood Pressure

The prognostic value of central systolic blood pressure has been established recently. At present, its noninvasive assessment is limited by the need of dedicated equipment and trained operators. Moreover, ambulatory and home blood pressure monitoring of central pressures are not feasible. An algorithm enabling conventional automated oscillometric blood pressure monitors to assess central systolic pressure could be of value. We compared central systolic pressure, calculated with a transfer-function like method (ARCSolver algorithm), using waveforms recorded with a regular oscillometric cuff suitable for ambulatory measurements, with simultaneous high-fidelity invasive recordings, and with noninvasive estimations using a validated device, operating with radial tonometry and a generalized transfer function. Both studies revealed a good agreement between the oscillometric cuff-based central systolic pressure and the comparator. In the invasive study, composed of 30 patients, mean difference between oscillometric cuff/ARCSolver-based and invasive central systolic pressures was 3.0 mm Hg (SD: 6.0 mm Hg) with invasive calibration of brachial waveforms and −3.0 mm Hg (SD: 9.5 mm Hg) with noninvasive calibration of brachial waveforms. Results were similar when the reference method (radial tonometry/transfer function) was compared with invasive measurements. In the noninvasive study, composed of 111 patients, mean difference between oscillometric cuff/ARCSolver–derived and radial tonometry/transfer function–derived central systolic pressures was −0.5 mm Hg (SD: 4.7 mm Hg). In conclusion, a novel transfer function-like algorithm, using brachial cuff-based waveform recordings, is suited to provide a realistic estimation of central systolic pressure.

Oscillometric estimation of aortic pulse wave velocity: comparison with intra-aortic catheter measurements.

ObjectivesRecently, a novel method to estimate aortic pulse wave velocity (aPWV) noninvasively from an oscillometric single brachial cuff waveform reading has been introduced. We investigated whether this new approach provides acceptable estimates of aPWV compared with intra-aortic catheter measurements. MethodsEstimated values of aPWV obtained from brachial cuff readings were compared with those obtained using an intra-aortic catheter in 120 patients (mean age 61.8±10.8 years) suspected for coronary artery disease undergoing cardiac catheterization. Differences between aPWV values obtained from the test device and those obtained from catheter measurements were estimated using Bland–Altman analysis. ResultsThe mean difference±SD between brachial cuff-derived values and intra-aortic values was 0.43±1.24 m/s. Comparison of aPWV measured by the two methods showed a significant linear correlation (Pearson’s R=0.81, P<0.0001). The mean difference for repeated oscillometric measurements of aPWV was 0.05 m/s, with 95% confidence interval limits from −0.47 to 0.57 m/s. ConclusionaPWV can be obtained using an oscillometric device with brachial cuffs with acceptable accuracy compared with intra-aortic readings.

Wave reflection quantification based on pressure waveforms alone-Methods, comparison, and clinical covariates

Within the last decade the quantification of pulse wave reflections mainly focused on measures of central aortic systolic pressure and its augmentation through reflections based on pulse wave analysis (PWA). A complementary approach is the wave separation analysis (WSA), which quantifies the total amount of arterial wave reflection considering both aortic pulse and flow waves. The aim of this work is the introduction and comparison of aortic blood flow models for WSA assessment. To evaluate the performance of the proposed modeling approaches (Windkessel, triangular and averaged flow), comparisons against Doppler measurements are made for 148 patients with preserved ejection fraction. Stepwise regression analysis between WSA and PWA parameters are performed to provide determinants of methodological differences. Against Doppler measurement mean difference and standard deviation of the amplitudes of the decomposed forward and backward pressure waves are comparable for Windkessel and averaged flow models. Stepwise regression analysis shows similar determinants between Doppler and Windkessel model only. The results indicate that the Windkessel method provides accurate estimates of wave reflection in subjects with preserved ejection fraction. The comparison with waveforms derived from Doppler ultrasound as well as recently proposed simple triangular and averaged flow waves showed that this approach may reduce variability and provide realistic results.

A modular platform for event recognition in smart homes

Ambient Assisted Living technologies try to integrate intelligent assistance-systems in elder peoples homes to maintain a high degree of independence, autonomy and dignity. To speed up the development process and to make the applications more adaptive and flexible to special user needs as well as to ensure compatibility throughout systems a common middleware with standardized interfaces is desirable. The integration of off-the-shelf sensor hardware is an important aspect to assure longterm applicability and interoperability. AIT Austrian Institute of Technology has developed a modular software platform providing services to enable independent living for elder people at home. The platform is based on state-of-the-art software development techniques like OSGi and Spring, which enable modularity and flexibility. The aspect of interoperability at the hardware level is considered by integrating standards from the two areas of medical informatics and home automation. A hardware abstraction module harmonizes data from different sensor networks in terms of a common data format. Based on sensor data, functionalities for the detection of specific events and situations in the AAL domain have been implemented using finite state machines and simple statistical methods.

Calculation of arterial characteristic impedance: a comparison using different blood flow models

Within the concept of pulse wave analysis, arterial pressure and flow curves over a whole cardiac cycle are analysed. Characteristic impedance is obtained as ratio of pressure to flow when waves are not influenced by reflections. The aim of this work is to evaluate the effects of different blood flow models on the determination of the characteristic impedance compared to flow curves gained from ultrasound measurements. Beside a simple triangular and an averaged flow, a new blood flow model based on Windkessel theory is used. In a study population of 148 patients for the evaluation of the different models, the characteristic impedance is calculated in the frequency domain. The results indicate that the characteristic impedance strongly depends on the accuracy of the used flow model. While the averaged and the ARCSolver flow provide good estimates for impedance, the triangular flow curve seems to be too simplistic for getting accurate values.

Smart Environments and Context-Awareness for Lifestyle Management in a Healthy Active Ageing Framework*

Health trends of elderly in Europe motivate the need for technological solutions aimed at preventing the main causes of morbidity and premature mortality. In this framework, the DOREMI project addresses three important causes of morbidity and mortality in the elderly by devising an ICT-based home care services for aging people to contrast cognitive decline, sedentariness and unhealthy dietary habits. In this paper, we present the general architecture of DOREMI, focusing on its aspects of human activity recognition and reasoning.

Human activity recognition using recurrent neural networks

Human activity recognition using smart home sensors is one of the bases of ubiquitous computing in smart environments and a topic undergoing intense research in the field of ambient assisted living. The increasingly large amount of data sets calls for machine learning methods. In this paper, we introduce a deep learning model that learns to classify human activities without using any prior knowledge. For this purpose, a Long Short Term Memory (LSTM) Recurrent Neural Network was applied to three real world smart home datasets. The results of these experiments show that the proposed approach outperforms the existing ones in terms of accuracy and performance.

Ambient Assisted Living Technologies from the Perspectives of Older People and Professionals

Ambient Assisted Living (AAL) and Ambient Intelligence technologies are providing support to older people in living an independent and confident life by developing innovative ICT-based products, services, and systems. Despite significant advancement in AAL technologies and smart systems, they have still not found the way into the nursing home of the older people. The reasons are manifold. On one hand, the development of such systems lack in addressing the requirements of the older people and caregivers of the organization and the other is the unwillingness of the older people to make use of assistive systems. A qualitative study was performed at a nursing home to understand the needs and requirements of the residents and caregivers and their perspectives about the existing AAL technologies.

Convolutional and Recurrent Neural Networks for Activity Recognition in Smart Environment

Convolutional Neural Networks (CNN) are very useful for fully automatic extraction of discriminative features from raw sensor data. This is an important problem in activity recognition, which is of enormous interest in ambient sensor environments due to its universality on various applications. Activity recognition in smart homes uses large amounts of time-series sensor data to infer daily living activities and to extract effective features from those activities, which is a challenging task. In this paper we demonstrate the use of the CNN and a comparison of results, which has been performed with Long Short Term Memory (LSTM), recurrent neural networks and other machine learning algorithms, including Naive Bayes, Hidden Markov Models, Hidden Semi-Markov Models and Conditional Random Fields. The experimental results on publicly available smart home datasets demonstrate that the performance of 1D-CNN is similar to LSTM and better than the other probabilistic models.

Original articles: Intelligent multichannel sensors for pulse wave analysis

Aortic pulse wave velocity is an independent predictive indicator for all cause mortality and cardiovascular morbidity. Unfortunately it is only invasively accessible and thus the A. carotis-A. femoralis pulse wave velocity (cfPWV) is recommended as a non-invasive substitute. This work presents a model based analysis method for the beat-to-beat online determination of an arbitrary, peripheral pulse transit time (PTT). The method is based on the recording of a three lead electrocardiography (ECG) and of pulse waves (PW) at a peripheral site such as the A. carotis by means of a multiple sensor array. The two modules for the signal acquisition work autonomously but time-wise simultaneously and transmit the data via a radio unit to the central processing unit. There the algorithms for the pulse transit time determination exploit these signals. In doing so the main focus is on an efficient implementation to assure real-time usability. The evaluation of the developed modules and algorithms was done in two separate trials. First the algorithms were tested offline against manual signal annotation using invasive data previously recorded to proof their accuracy. The resulting mean differences in PTT for pulse waves assessed at the aortic root and the aortic bifurcation are 2.86ms (4.72ms SD) and 2.00ms (2.28ms SD). To evaluate the whole system integrity in a second step online measurements on probands were carried out and compared to data from literature. The trials resulted in a mean PTT of 174.6ms (17.7ms SD) for the A. radialis and of 81.9ms (13.2ms SD) for the A. carotis. The results suggest that the method may be useful and deployable at general practitioners (GP) and in ambulatory care of (chronic) cardiovascular diseases.