Remote Monitoring of Positive Airway Pressure Data: Challenges, Pitfalls, and Strategies to Consider for Optimal Data Science Applications.

Over recent years, positive airway pressure (PAP) remote monitoring has transformed the management of OSA and produced a large amount of data. Accumulated PAP data provide valuable and objective information regarding patient treatment adherence and efficiency. However, the majority of studies that have analyzed longitudinal PAP remote monitoring have summarized data trajectories in static and simplistic metrics for PAP adherence and the residual apnea-hypopnea index by the use of mean or median values. The aims of this article are to suggest directions for improving data cleaning and processing and to address major concerns for the following data science applications: (1) conditions for residual apnea-hypopnea index reliability, (2) lack of standardization of indicators provided by different PAP models, (3) missing values, and (4) consideration of treatment interruptions. To allow fair comparison among studies and to avoid biases in computation, PAP data processing and management should be conducted rigorously with these points in mind. PAP remote monitoring data contain a wealth of information that currently is underused in the field of sleep research. Improving the quality and standardizing data handling could facilitate data sharing among specialists worldwide and enable artificial intelligence strategies to be applied in the field of sleep apnea.

Bayesian Structural Time Series With Synthetic Controls for Evaluating the Impact of Mask Changes in Residual Apnea-Hypopnea Index Telemonitoring Data.

OBJECTIVE: In obstructive sleep apnea patients on continuous positive airway pressure (CPAP) treatment there is growing evidence for a significant impact of the type of mask on the residual apnea-hypopnea index (rAHI). Here, we propose a method for automatically classifying the impact of mask changes on rAHI. METHODS: From a CPAP telemonitoring database of 3,581 patients, an interrupted time series design was applied to rAHI time series at a patient level to compare the observed rAHI after a mask-change with what would have occurred without the mask-change. rAHI time series before mask changes were modelled using different approaches. Mask changes were classified as: no effect, harmful, beneficial. The best model was chosen based on goodness-of-fit metrics and comparison with blinded classification by an experienced respiratory physician. RESULTS: Bayesian structural time series with synthetic controls was the best approach in terms of agreement with the physician.s classification, with an accuracy of 0.79. Changes from nasal to facial mask were more often harmful than beneficial: 13.4% vs 7.6% (p-value < 0.05), with a clinically relevant increase in average rAHI greater than 8 events/hour in 4.6% of cases. Changes from facial to nasal mask were less often harmful: 6.0% vs 11.4% (p-value < 0.05). CONCLUSION: We propose an end-to-end method to automatically classify the impact of mask changes over fourteen days after a switchover. SIGNIFICANCE: The proposed automated analysis of the impact of changes in health device settings or accessories presents a novel tool to include in remote monitoring platforms for raising alerts after harmful interventions.

Hidden Markov model segmentation to demarcate trajectories of residual apnoea-hypopnoea index in CPAP-treated sleep apnoea patients to personalize follow-up and prevent treatment failure.

BACKGROUND: Continuous positive airway pressure (CPAP), the reference treatment for obstructive sleep apnoea (OSA), is used by millions of individuals worldwide with remote telemonitoring providing daily information on CPAP usage and efficacy, a currently underused resource. Here, we aimed to implement data science methods to provide tools for personalizing follow-up and preventing treatment failure. METHODS: We analysed telemonitoring data from adults prescribed CPAP treatment. Our primary objective was to use Hidden Markov models (HMMs) to identify the underlying state of treatment efficacy and enable early detection of deterioration. Secondary goals were to identify clusters of rAHI trajectories which need distinct therapeutic strategies. RESULTS: From telemonitoring records of 2860 CPAP-treated patients (age: 66.31 ± 12.92 years, 69.9% male), HMM estimated three states differing in variability within a given state and probability of shifting from one state to another. The daily inferred state informs on the need for a personalized action, while the sequence of states is a predictive indicator of treatment failure. Six clusters of rAHI trajectories were identified ranging from well-controlled patients (cluster 0: 669 (23%); mean rAHI 0.58 ± 0.59 events/h) to the most unstable (cluster 5: 470 (16%); mean rAHI 9.62 ± 5.62 events/h). CPAP adherence was 30 min higher in cluster 0 compared to clusters 4 and 5 (P value < 0.01). CONCLUSION: This new approach based on HMM might constitute the backbone for deployment of patient-centred CPAP management improving the personalized interpretation of telemonitoring data, identifying individuals for targeted therapy and preventing treatment failure or abandonment. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s13167-021-00264-z.

Apnea-hypopnea index supplied by CPAP devices: time for standardization?

BACKGROUND/OBJECTIVE: For obstructive sleep apnea (OSA) patients on continuous positive airway pressure (CPAP) treatment, the apnea-hypopnea index (AHI) is a key measure of treatment efficacy. However, the residual AHI is CPAP brand specific. Here, we studied changes in residual AHI in patients who used two different brands over their treatment history. PATIENTS/METHODS: Using our CPAP telemonitoring database of 3102 patients, we compared the residual AHI of 69 patients before and after change in their CPAP device. RESULTS: A paired Wilcoxon signed-rank test revealed a significant difference between brands in the reported AHI, which might be clinically misleading. CONCLUSIONS: These findings suggest that physicians should be alerted to the differences between brands and learned societies should push for standardization of AHI reporting.

An exoskeleton controlled by an epidural wireless brain-machine interface in a tetraplegic patient: a proof-of-concept demonstration.

BACKGROUND: Approximately 20% of traumatic cervical spinal cord injuries result in tetraplegia. Neuroprosthetics are being developed to manage this condition and thus improve the lives of patients. We aimed to test the feasibility of a semi-invasive technique that uses brain signals to drive an exoskeleton. METHODS: We recruited two participants at Clinatec research centre, associated with Grenoble University Hospital, Grenoble, France, into our ongoing clinical trial. Inclusion criteria were age 18-45 years, stability of neurological deficits, a need for additional mobility expressed by the patient, ambulatory or hospitalised monitoring, registration in the French social security system, and signed informed consent. The exclusion criteria were previous brain surgery, anticoagulant treatments, neuropsychological sequelae, depression, substance dependence or misuse, and contraindications to magnetoencephalography (MEG), EEG, or MRI. One participant was excluded because of a technical problem with the implants. The remaining participant was a 28-year-old man, who had tetraplegia following a C4-C5 spinal cord injury. Two bilateral wireless epidural recorders, each with 64 electrodes, were implanted over the upper limb sensorimotor areas of the brain. Epidural electrocorticographic (ECoG) signals were processed online by an adaptive decoding algorithm to send commands to effectors (virtual avatar or exoskeleton). Throughout the 24 months of the study, the patient did various mental tasks to progressively increase the number of degrees of freedom. FINDINGS: Between June 12, 2017, and July 21, 2019, the patient cortically controlled a programme that simulated walking and made bimanual, multi-joint, upper-limb movements with eight degrees of freedom during various reach-and-touch tasks and wrist rotations, using a virtual avatar at home (64·0% [SD 5·1] success) or an exoskeleton in the laboratory (70·9% [11·6] success). Compared with microelectrodes, epidural ECoG is semi-invasive and has similar efficiency. The decoding models were reusable for up to approximately 7 weeks without recalibration. INTERPRETATION: These results showed long-term (24-month) activation of a four-limb neuroprosthetic exoskeleton by a complete brain-machine interface system using continuous, online epidural ECoG to decode brain activity in a tetraplegic patient. Up to eight degrees of freedom could be simultaneously controlled using a unique model, which was reusable without recalibration for up to about 7 weeks. FUNDING: French Atomic Energy Commission, French Ministry of Health, Edmond J Safra Philanthropic Foundation, Fondation Motrice, Fondation Nanosciences, Institut Carnot, Fonds de Dotation Clinatec.

Data-Driven Transducer Design and Identification for Internally-Paced Motor Brain Computer Interfaces: A Review.

Brain-Computer Interfaces (BCIs) are systems that establish a direct communication pathway between the users' brain activity and external effectors. They offer the potential to improve the quality of life of motor-impaired patients. Motor BCIs aim to permit severely motor-impaired users to regain limb mobility by controlling orthoses or prostheses. In particular, motor BCI systems benefit patients if the decoded actions reflect the users' intentions with an accuracy that enables them to efficiently interact with their environment. One of the main challenges of BCI systems is to adapt the BCI's signal translation blocks to the user to reach a high decoding accuracy. This paper will review the literature of data-driven and user-specific transducer design and identification approaches and it focuses on internally-paced motor BCIs. In particular, continuous kinematic biomimetic and mental-task decoders are reviewed. Furthermore, static and dynamic decoding approaches, linear and non-linear decoding, offline and real-time identification algorithms are considered. The current progress and challenges related to the design of clinical-compatible motor BCI transducers are additionally discussed.

Switching Markov decoders for asynchronous trajectory reconstruction from ECoG signals in monkeys for BCI applications.

Brain-Computer Interfaces (BCIs) are systems which translate brain neural activity into commands for external devices. BCI users generally alternate between No-Control (NC) and Intentional Control (IC) periods. NC/IC discrimination is crucial for clinical BCIs, particularly when they provide neural control over complex effectors such as exoskeletons. Numerous BCI decoders focus on the estimation of continuously-valued limb trajectories from neural signals. The integration of NC support into continuous decoders is investigated in the present article. Most discrete/continuous BCI hybrid decoders rely on static state models which don't exploit the dynamic of NC/IC state succession. A hybrid decoder, referred to as Markov Switching Linear Model (MSLM), is proposed in the present article. The MSLM assumes that the NC/IC state sequence is generated by a first-order Markov chain, and performs dynamic NC/IC state detection. Linear continuous movement models are probabilistically combined using the NC and IC state posterior probabilities yielded by the state decoder. The proposed decoder is evaluated for the task of asynchronous wrist position decoding from high dimensional space-time-frequency ElectroCorticoGraphic (ECoG) features in monkeys. The MSLM is compared with another dynamic hybrid decoder proposed in the literature, namely a Switching Kalman Filter (SKF). A comparison is additionally drawn with a Wiener filter decoder which infers NC states by thresholding trajectory estimates. The MSLM decoder is found to outperform both the SKF and the thresholded Wiener filter decoder in terms of False Positive Ratio and NC/IC state detection error. It additionally surpasses the SKF with respect to the Pearson Correlation Coefficient and Root Mean Squared Error between true and estimated continuous trajectories.

Time-division multiplexing for myoelectric closed-loop control using electrotactile feedback.

BACKGROUND: Restoring sensory feedback in myoelectric prostheses is still an open challenge. Closing the loop might lead to a more effective utilization and better integration of these systems into the body scheme of the user. Electrotactile stimulation can be employed to transmit the feedback information to the user, but it represents a strong interference to the recording of the myoelectric signals that are used for control. Time-division multiplexing (TDM) can be applied to avoid this interference by performing the stimulation and recording in dedicated, non-overlapping time windows. METHODS: A closed-loop compensatory tracking task with myocontrol and electrotactile stimulation was used to investigate how the duration of the feedback window (FW) influences the ability to perceive the feedback information and react with an appropriate control action. Nine subjects performed eight trials with continuous recording and contralateral feedback (CONT-CLT) and TDM with ispilateral stimulation and recording using the FW of 40 ms (TDM40), 100 ms (TDM100) and 300 ms (TDM300). The tracking quality was evaluated by comparing the reference and generated trajectories using cross-correlation coefficient (CCCOEF), time delay, root mean square tracking error, and the amount of overshoot. RESULTS: The control performance in CONT-CLT was the best in all the outcome measures. The overall worst performance was obtained using TDM with the shortest FW (TDM40). There was no significant difference between TDM100 and TDM300, and the quality of tracking in these two conditions was high (CCCOEF ~ 0.95). The results demonstrated that FW duration is indeed an important parameter in TDM, which appears to have an optimal value. Among the tested cases, the FW duration of 100 ms seems to be the best trade-off between the quality of perception and a limited command update rate. CONCLUSIONS: This study represents the first systematic evaluation of a TDM-based approach for closing the loop using electrotactile feedback in myoelectric systems. The overall conclusion is that TDM is a feasible and attractive method for closed-loop myocontrol, since it is easy to implement (software-only solution), has limited impact on the performance when using proper FW duration, and might decrease habituation due to burst-like stimulation delivery.