Signal quality measures on pulse oximetry and blood pressure signals acquired from self-measurement in a home environment

Enabling Guidelines for the Adoption of eHealth Solutions: Scoping Review

BACKGROUND

Globally, public health care is under increasing pressure, an economic burden currently amplified by the COVID-19 outbreak. With the recognition that universal health coverage improves the health of a population and reduces health inequalities, universal health coverage has been acknowledged as a priority goal. To meet the global needs in a population with increased chronic illness and longer life expectancy, the health care system is in dire need of new, emerging technologies. eHealth solutions as a method of delivery may have an impact on quality of care and health care costs. As such, it is important to study methods previously used to avoid suboptimal implementation and promote general guidelines to further develop eHealth solutions.

OBJECTIVE

This study aims to explore and thematically categorize a selected representation of early phase studies on eHealth technologies, focusing on papers that are under development or undergoing testing. Further, we want to assess enablers and barriers in terms of usability, scaling, and data management of eHealth implementation. The aim of this study to explore early development phase and feasibility studies was an intentional effort to provide applicable guidelines for evaluation at different stages of implementation.

METHODS

A structured search was performed in PubMed, MEDLINE, and Cochrane to identify and provide insight in current eHealth technology and methodology under development and gain insight in the future potential of eHealth technologies.

RESULTS

In total, 27 articles were included in this review. The clinical studies were categorized thematically by illness comparing 4 technology types deemed relevant: apps/web-based technology, sensor technology, virtual reality, and television. All eHealth assessment and implementation studies were categorized by their focus point: usability, scaling, or data management. Studies assessing the effect of eHealth were divided into feasibility studies, qualitative studies, and heuristic assessments. Studies focusing on usability (16/27) mainly addressed user involvement and learning curve in the adoption of eHealth, while the majority of scaling studies (6/27) focused on strategic and organizational aspects of upscaling eHealth solutions. Studies focusing on data management (5/27) addressed data processing and data sensitivity in adoption and diffusion of eHealth. Efficient processing of data in a secure manner, as well as user involvement and feedback, both throughout small studies and during upscaling, were the important enablers considered for successful implementation of eHealth.

CONCLUSIONS

eHealth interventions have considerable potential to improve lifestyle changes and adherence to treatment recommendations. To promote efficient implementation and scaling, user involvement to promote user-friendliness, secure and adaptable data management, and strategical considerations needs to be addressed early in the development process. eHealth should be assessed during its development into health services. The wide variation in interventions and methodology makes comparison of the results challenging and calls for standardization of methods.

On-Device Reliability Assessment and Prediction of Missing Photoplethysmographic Data Using Deep Neural Networks

Photoplethysmographic (PPG) measurements from ambulatory subjects may suffer from unreliability due to body movements and missing data segments due to loosening of sensor. This paper describes an on-device reliability assessment from PPG measurements using a stack denoising autoencoder (SDAE) and multilayer perceptron neural network (MLPNN). The missing segments were predicted by a personalized convolutional neural network (CNN) and long-short term memory (LSTM) model using a short history of the same channel data. Forty sets of volunteers' data, consisting of equal share of healthy and cardiovascular subjects were used for validation and testing. The PPG reliability assessment model (PRAM) achieved over 95% accuracy for correctly identifying acceptable PPG beats out of total 5000 using expert annotated data. Disagreement with experts' annotation was nearly 3.5%. The missing segment prediction model (MSPM) achieved a root mean square error (RMSE) of 0.22, and mean absolute error (MAE) of 0.11 for 40 missing beats prediction using only four beat history from the same channel PPG. The two models were integrated in a standalone device based on quad-core ARM Cortex-A53, 1.2 GHz, with 1 GB RAM, with 130 MB memory requirement and latency ∼0.35 s per beat prediction with a 30 s frame. The present method also provides improved performance with published works on PPG quality assessment and missing data prediction using two public datasets, CinC and MIMIC-II under PhysioNet.

A Photoplethysmographic Signal Isolated From an Additive Motion Artifact by Frequency Translation

Acquiring a precise percentage of oxygen saturation (SpO2) from a finger-probe pulse oximeter is dependent on both artifact-free red and infrared photoplethysmoghaphic (PPG) signals. Nonetheless, in real-life situations, these PPG signals are corrupted by a motion artifact (MA) signal that is generated from either finger or hand movement. To resolve this MA interference, the cause of the adulteration of PPG signals by the MA signal is examined. The MA signal is found to behave like an additive noise. Additionally, the frequency responses of the MA and PPG signals show that these signals are in the same frequency band. Hence, instead of direct current, a sinusoidal wave alternating current is proposed to drive an LED source in order to shift the PPG frequency band away from the MA frequency band. Experimentally, a commercial finger-probe pulse oximeter is employed. To determine the performance of the presented scheme, the resulting PPG signals are compared with those from employing the old-fashioned LED-driving method. In addition, the accuracy is verified by computing the SpO2 value. The results reveal that the proposed approach successfully retains the fundamental morphologies of the PPG structures when motion occurs. Moreover, the calculated SpO2 values from the proposed technique provide an average error of approximately 1.4%, whereas the conventional method yields a mean error approximately 4.2%.

Validation of overnight oximetry to diagnose patients with moderate to severe obstructive sleep apnea

Background

Polysomnography (PSG) is treated as the gold standard for diagnosing obstructive sleep apnea (OSA). However, it is labor-intensive, time-consuming, and expensive. This study evaluates validity of overnight pulse oximetry as a diagnostic tool for moderate to severe OSA patients.

Methods

A total of 699 patients with possible OSA were recruited for overnight oximetry and PSG examination at the Sleep Center of a University Hospital from Jan. 2004 to Dec. 2005. By excluding 23 patients with poor oximetry recording, poor EEG signals, or respiratory artifacts resulting in a total recording time less than 3 hours; 12 patients with total sleeping time (TST) less than 1 hour, possibly because of insomnia; and 48 patients whose ages less than 20 or more than 85 years old, data of 616 patients were used for further study. By further considering 76 patients with TST < 4 h, a group of 540 patients with TST ≥ 4 h was used to study the effect of insufficient sleeping time. Alice 4 PSG recorder (Respironics Inc., USA) was used to monitor patients with suspected OSA and to record their PSG data. After statistical analysis and feature selection, models built based on support vector machine (SVM) were then used to diagnose moderate and moderate to severe OSA patients with a threshold of AHI = 30 and AHI = 15, respectively.

Results

The SVM models designed based on the oxyhemoglobin desaturation index (ODI) derived from oximetry measurements provided an accuracy of 90.42-90.55%, a sensitivity of 89.36-89.87%, a specificity of 91.08-93.05%, and an area under ROC curve (AUC) of 0.953-0.957 for the diagnosis of severe OSA patients; as well as achieved an accuracy of 87.33-87.77%, a sensitivity of 87.71-88.53%, a specificity of 86.38-86.56%, and an AUC of 0.921-0.924 for the diagnosis of moderate to severe OSA patients. The predictive outcome of ODI to diagnose severe OSA patients is better than to diagnose moderate to severe OSA patients.

Conclusions

Overnight pulse oximetry provides satisfactory diagnostic performance in detecting severe OSA patients. Home-styled oximetry may be a tool for severe OSA diagnosis.