Resource-Aware Time Series Imaging Classification for Wireless Link Layer Anomalies.

The number of end devices that use the last-mile wireless connectivity is dramatically increasing with the rise of smart infrastructures and requires reliable functioning to support smooth and efficient business processes. To efficiently manage such massive wireless networks, more advanced and accurate network monitoring and malfunction detection solutions are required. In this article, we perform a first-time analysis of image-based representation techniques for wireless anomaly detection using recurrence plots (RPs) and Gramian angular fields and propose a new deep learning architecture enabling accurate anomaly detection. We elaborate on the design considerations for developing a resource-aware architecture and propose a new model using time series to image transformation using RPs. We show that the proposed model: 1) outperforms the one based on Gramian angular fields by up to 14% points; 2) outperforms classical ML models using dynamic time warping by up to 24% points; 3) outperforms or performs on par with mainstream architectures, such as AlexNet and VGG11 while having their weights and up to ≈8% of their computational complexity; and d) outperforms the state of the art in the respective application area by up to 55% points. Finally, we also explain on randomly chosen examples how the classifier takes decisions.

Design of a secure remote management module for a software-operated medical device.

Software-based medical devices need to be maintained throughout their entire life cycle. The efficiency of after-sales maintenance can be improved by managing medical systems remotely. This paper presents how to design the remote access function extensions in order to prevent risks imposed by uncontrolled remote access. A thorough analysis of standards and legislation requirements regarding safe operation and risk management of medical devices is presented. Based on the formal requirements, a multi-layer machine design solution is proposed that eliminates remote connectivity risks by strict separation of regular device functionalities from remote management service, deploys encrypted communication links and uses digital signatures to prevent mishandling of software images. The proposed system may also be used as an efficient version update of the existing medical device designs.

Detection of peripheral arterial disease with an improved automated device: comparison of a new oscillometric device and the standard Doppler method.

In occidental countries, peripheral arterial disease (PAD) is an important health issue; however, most subjects are asymptomatic (~50%) and therefore undiagnosed and untreated. Current guidelines recommend screening for PAD in primary care setting using ankle brachial index (ABI) in all patients with cardiovascular risks. This is, however, not performed strictly because the standard Doppler method is cumbersome and time-consuming. Here, we evaluate the accuracy and reproducibility of ABI measurements obtained by an improved automated oscillometric device, the MESI ABPI MD(®) device, and the standard Doppler method. ABI was measured in random order in a general practice with Doppler probes by two operators separately (ABI_dop) and twice with the MESI ABPI MD device (ABI_mesi). ABI_dop was calculated dividing the highest systolic blood pressure from both tibial and dorsalis pedis arteries by the highest systolic blood pressure of both brachial arteries. ABI_mesi was obtained automatically with simultaneous measurements on three extremities. According to ABI_dop, PAD was present in 10% of the 136 screened subjects (68.2±7.4 years). Interoperator coefficient of variation was 5.5% for ABI_dop, while the intrasubject coefficient of variation for ABI_mesi was 3.0%. ABI_mesi was correlated with ABI_dop (R=0.61, P<0.0001). The difference between the two techniques was 0.06±0.14 with ABI_mesi providing slightly higher values (P<0.0001) and negligible bias across the range (R=0.19, P<0.0001). Therefore, ABI_mesi ≤1 had a sensitivity of 85% and specificity of 96% to detect ABI_dop ≤0.9 and hence PAD. Doppler measurements took seven times longer than MESI ABPI MD measurements to be performed. In conclusion, MESI improved automated oscillometric method and offered a faster and repeatable measurement of ABI with only a small, clinically irrelevant overestimation of ABI value. The tested MESI ABPI MD-improved oscillometric system can be used as a screening tool for patients in general practice and would enable family doctors to comply with current guidelines for PAD.

Telemedicine in the blood transfusion laboratory: remote interpretation of pre-transfusion tests.

We have developed a telemedicine system for blood transfusion work, to supply the local hospital laboratory with an expert opinion from the central reference laboratory. The telemedicine system allows remote inspection and interpretation of pre-transfusion tests, which are performed by ID-cards (micro-tube gel technology). The system was installed at three blood transfusion laboratories in Slovenia, approximately 70 km apart. Validation of the telemedicine system was performed using 99 clinical cases selected randomly from routine work. Two groups of immunohaematology specialists participated. Group A (n = 8) performed the read-out of the pre-transfusion tests on ID-cards by using the telemedicine system. Group B (n = 2) then read the ID-cards independently using the standard visual method. All 98 final interpretations which were recorded using the telemedicine system were correct. We recorded 591 micro-tube read-outs of agglutination strength using the telemedicine system, of which 582 were correct. For comparison, we recorded 591 micro-tube read-outs using the standard visual method, of which 582 were correct. The validation proved that the telemedicine system was suitable for operational use.