Electromyography (EMG), the measurement of electrical muscle activity, is used in a variety of applications, including myoelectric upper-limb prostheses, which help amputees to regain independence and a higher quality of life. The state-of-the-art sensors in prostheses have a conductive connection to the skin and are therefore sensitive to sweat and require preparation of the skin. They are applied with some pressure to ensure a conductive connection, which may result in pressure marks and can be problematic for patients with circulatory disorders, who constitute a major group of amputees. Due to their insulating layer between skin and sensor area, capacitive sensors are insensitive to the skin condition, they require neither conductive connection to the skin nor electrolytic paste or skin preparation. Here, we describe a highly stable, low-power capacitive EMG measurement set-up that is suitable for real-world application. Various flexible multi-layer sensor set-ups made of copper and insulating foils, flex print and textiles were compared. These flexible sensor set-ups adapt to the anatomy of the human forearm, therefore they provide high wearing comfort and ensure stability against motion artifacts. The influence of the materials used in the sensor set-up on the magnitude of the coupled signal was demonstrated based on both theoretical analysis and measurement.The amplifier circuit was optimized for high signal quality, low power consumption and mobile application. Different shielding and guarding concepts were compared, leading to high SNR.

Many previous studies claim to have developed machine learning models that diagnose COVID-19 from blood tests. However, we hypothesize that changes in the underlying distribution of the data, so called domain shifts, affect the predictive performance and reliability and are a reason for the failure of such machine learning models in clinical application. Domain shifts can be caused, e.g., by changes in the disease prevalence (spreading or tested population), by refined RT-PCR testing procedures (way of taking samples, laboratory procedures), or by virus mutations. Therefore, machine learning models for diagnosing COVID-19 or other diseases may not be reliable and degrade in performance over time. We investigate whether domain shifts are present in COVID-19 datasets and how they affect machine learning methods. We further set out to estimate the mortality risk based on routinely acquired blood tests in a hospital setting throughout pandemics and under domain shifts. We reveal domain shifts by evaluating the models on a large-scale dataset with different assessment strategies, such as temporal validation. We present the novel finding that domain shifts strongly affect machine learning models for COVID-19 diagnosis and deteriorate their predictive performance and credibility. Therefore, frequent re-training and re-assessment are indispensable for robust models enabling clinical utility.

Electronic health records (EHRs) have been successfully used in data science and machine learning projects. However, most of these data are collected for clinical use rather than for retrospective analysis. This means that researchers typically face many different issues when attempting to access and prepare the data for secondary use. We aimed to investigate how raw EHRs can be accessed and prepared in retrospective data science projects in a disciplined, effective, and efficient way. We report our experience and findings from a large-scale data science project analyzing routinely acquired retrospective data from the Kepler University Hospital in Linz, Austria. The project involved data collection from more than 150,000 patients over a period of 10 years. It included diverse data modalities, such as static demographic data, irregularly acquired laboratory test results, regularly sampled vital signs, and high-frequency physiological waveform signals. Raw medical data can be corrupted in many unexpected ways that demand thorough manual inspection and highly individualized data cleaning solutions. We present a general data preparation workflow, which was shaped in the course of our project and consists of the following 7 steps: obtain a rough overview of the available EHR data, define clinically meaningful labels for supervised learning, extract relevant data from the hospital’s data warehouses, match data extracted from different sources, deidentify them, detect errors and inconsistencies therein through a careful exploratory analysis, and implement a suitable data processing pipeline in actual code. Only few of the data preparation issues encountered in our project were addressed by generic medical data preprocessing tools that have been proposed recently. Instead, highly individualized solutions for the specific data used in one’s own research seem inevitable. We believe that the proposed workflow can serve as a guidance for practitioners, helping them to identify and address potential problems early and avoid some common pitfalls.

Myoelectric prostheses help amputees to regain independence and a higher quality of life. These prostheses are controlled by state-of-the-art electromyography sensors, which use a conductive connection to the skin and are therefore sensitive to sweat. They are applied with some pressure to ensure a conductive connection, which may result in pressure marks and can be problematic for patients with circulatory disorders, who constitute a major group of amputees. Here, we present ultra-low-power digital signal processing algorithms for an insulated EMG sensor which couples the EMG signal capacitively. These sensors require neither conductive connection to the skin nor electrolytic paste or skin preparation. Capacitive sensors allow straightforward application. However, they make a sophisticated signal amplification and noise suppression necessary. A low-cost sensor has been developed for real-time myoelectric prostheses control. The major hurdles in measuring the EMG are movement artifacts and external noise. We designed various digital filters to attenuate this noise. Optimal system setup and filter parameters for the trade-off between attenuation of this noise and sufficient EMG signal power for high signal quality were investigated. Additionally, an algorithm for movement artifact suppression, enabling robust application in real-world environments, is presented. The algorithms, which require minimal calculation resources and memory, are implemented on an ultra-low-power microcontroller.

Myoelectric prostheses help amputees to regain independence and a higher quality of life. These prostheses are controlled by electromyography, which measures an electrical signal at the skin surface during muscle contractions. In this contribution, the electromyography is measured with innovative flexible insulated sensors, which separate the skin and the sensor area by a dielectric layer. Electromyography sensors, and biosignal sensors in general, are striving for higher robustness against motion artifacts, which are a major obstacle in real-world environment. The motion artifact suppression algorithms presented in this article, prevent the activation of the prosthesis drive during artifacts, thereby achieving a substantial performance boost. These algorithms classify the signal into muscle contractions and artifacts. Therefore, new time domain features, such as Mean Crossing Rate are introduced and well-established time domain features (e.g., Zero-Crossing Rate, Slope Sign Change) are modified and implemented. Various artificial intelligence models, which require low calculation resources for an application in a wearable device, were investigated. These models are neural networks, recurrent neural networks, decision trees and logistic regressions. Although these models are designed for a low-power real-time embedded system, high accuracies in discriminating artifacts to contractions of up to 99.9% are achieved. The models were implemented and trained for fast response leading to a high performance in real-world environment. For highest accuracies, recurrent neural networks are suggested and for minimum runtime (0.99–1.15 $\mathsf{\mu }$ s), decision trees are preferred.