Artifacts classification and apnea events detection in neck photoplethysmography signals

A novel computational signal processing framework towards multimodal vital signs extraction using neck-worn wearable devices

Pulse rate (PR) and respiratory rate (RR) are two of the most important vital signs. Monitoring them would benefit from easy-to-use technologies. Hence, wearable devices would, in principle, be ideal candidates for such systems. The neck, although highly susceptible to artifacts, presents an attractive location for a diverse pool of physiological biomarkers monitoring purposes such as airflow sensing in a non-obstructive manner. This paper presents a methodology for PR and RR estimation using photoplethysmography (PPG) and accelerometry (Acc) sensors placed on the neck. Neck PPG and Acc signals were recorded from 22 healthy participants for RR estimation, where the resting subjects performed guided breathing following a visual metronome. Neck PPG signals were obtained from 16 healthy participants who breathed through an altitude generator machine in order to acquire a wider range of PR readings while at rest. The proposed methodology was able to provide rate estimates via a combination of recursive FFT-based dominance scoring coupled with an exponentially weighted moving average (EWMA)-driven aggregation scheme. The recursion aimed at bypassing sudden intra-window amplitude deviations caused by momentary artifacts, while the EWMA-based aggregation was utilized for handling inter-window artifact-induced deviations. To further improve estimation stability and confidence, estimates were calculated in the form of rate bands taking into account the relevant clinically acceptable error margins, and results when considering rate values and rate bands are presented and discussed. The framework was able to achieve an overall pulse rate value accuracy of 93.67 ± 7.64 % within the clinically acceptable ± 5 BPM with reference to the gold-standard reference devices while providing an overall respiratory rate value accuracy within the clinically appropriate ± 3 BrPM of 94.94 ± 3.56 % with reference to the guiding visual metronome, and 88.4 ± 7.63 % with respect to the gold-standard reference device. The proposed methodology achieves acceptable PR and RR estimation capabilities, even when signals are acquired from an unusual location such as the neck. This work introduces novel ideas that can lead to the development of medical device outputs for PR and RR monitoring, especially capitalizing on the advantages of the neck as a multi-modal physiological monitoring location.

A comparative study in class imbalance mitigation when working with physiological signals

Class imbalance is a common challenge that is often faced when dealing with classification tasks aiming to detect medical events that are particularly infrequent. Apnoea is an example of such events. This challenge can however be mitigated using class rebalancing algorithms. This work investigated 10 widely used data-level class imbalance mitigation methods aiming towards building a random forest (RF) model that attempts to detect apnoea events from photoplethysmography (PPG) signals acquired from the neck. Those methods are random undersampling (RandUS), random oversampling (RandOS), condensed nearest-neighbors (CNNUS), edited nearest-neighbors (ENNUS), Tomek's links (TomekUS), synthetic minority oversampling technique (SMOTE), Borderline-SMOTE (BLSMOTE), adaptive synthetic oversampling (ADASYN), SMOTE with TomekUS (SMOTETomek) and SMOTE with ENNUS (SMOTEENN). Feature-space transformation using PCA and KernelPCA was also examined as a potential way of providing better representations of the data for the class rebalancing methods to operate. This work showed that RandUS is the best option for improving the sensitivity score (up to 11%). However, it could hinder the overall accuracy due to the reduced amount of training data. On the other hand, augmenting the data with new artificial data points was shown to be a non-trivial task that needs further development, especially in the presence of subject dependencies, as was the case in this work.

Quantifying the Suitability of Biosignals Acquired During Surgery for Multimodal Analysis

Goal: Recently, large datasets of biosignals acquired during surgery became available. As they offer multiple physiological signals measured in parallel, multimodal analysis - which involves their joint analysis - can be conducted and could provide deeper insights than unimodal analysis based on a single signal. However, it is unclear what percentage of intraoperatively acquired data is suitable for multimodal analysis. Due to the large amount of data, manual inspection and labelling into suitable and unsuitable segments are not feasible. Nevertheless, multimodal analysis is performed successfully in sleep studies since many years as their signals have proven suitable. Hence, this study evaluates the suitability to perform multimodal analysis on a surgery dataset (VitalDB) using a multi-center sleep dataset (SIESTA) as reference. Methods: We applied widely known algorithms entitled "signal quality indicators" to the common biosignals in both datasets, namely electrocardiography, electroencephalography, and respiratory signals split in segments of 10 s duration. As there are no multimodal methods available, we used only unimodal signal quality indicators. In case, all three signals were determined as being adequate by the indicators, we assumed that the whole signal segment was suitable for multimodal analysis. Results: 82% of SIESTA and 72% of VitalDB are suitable for multimodal analysis. Unsuitable signal segments exhibit constant or physiologically unreasonable values. Histogram examination indicated similar signal quality distributions between the datasets, albeit with potential statistical biases due to different measurement setups. Conclusions: The majority of data within VitalDB is suitable for multimodal analysis.