Onset Detection Research Articles

Predicting Rolling Element Bearings’ Deterioration Vibration Trend based on Limited Historical Data for a Desired Confidence Level using Machine Learning Algorithms

Predicting the deterioration trend of bearings and determining the optimal replacement time is crucial for maintenance teams to effectively manage consumable parts, enhance equipment reliability, and prevent potential breakdowns. Given the common practice of periodic condition monitoring based on vibrations in large industries, utilizing this data to forecast future vibration trends is a key focus of this research. Considering the limited availability of data in industrial settings, it is essential for the model used to be capable of working with minimal data. To address these challenges, a run-to-failure test has been conducted on an industrial bearing sample, with its vibration data recorded. Acknowledging the impact of operational variations and uncertainties in vibration behavior even among identical bearings, leveraging the bearing’s vibration history for predictive modeling has been deemed critical. This study proceeds in three distinct phases: selecting the key characteristic for predicting its deterioration trend, identifying the feature to pinpoint the failure onset, and determining the most suitable model for forecasting future bearing vibration states. RMS has emerged as the optimal characteristic for trend prediction, while Peak and Kurtosis have been identified as effective indicators for failure onset detection. The selection of the deterioration estimation model has involved comparing the outcomes of SVR + Bootstrapping and RVR models under various limited data scenarios. Ultimately, the RVR method has proved superior, requiring just three data points for estimation and delivering results with a confidence level leading to an accuracy of approximately 94% for the analyzed bearing.

Revealing Relationship Between In Situ Impedance and Lithium Plating Onset Based on Lithium–Graphite Half-Cells

Lithium plating may occur during charging, especially at high rates or overcharging conditions for lithium-ion batteries (LIBs), which would cause battery capacity degradation and even trigger thermal runaway. Thus, it is essential to detect lithium plating onset during the charging processes. Electrochemical impedance can reveal the dynamic electrode properties of the battery, which is promising for use in battery management systems for the online detection of lithium plating onset. In this article, the impedance at 1 Hz is measured during the over-discharge and fast discharge processes using lithium–graphite half-cells. For half-cells, the variation in graphite electrode potential vs. Li/Li+ during discharging is directly recorded. An equivalent circuit model is proposed and adopted to estimate the real lithium plating reaction overpotential, which is deemed the thermodynamic indicator of lithium plating and is used as validation for the detection of lithium plating onset. Through the auxiliary validation of the estimation of lithium plating overpotential and the shape of incremental capacity curves, the relationship between impedance changes at specific frequency and the lithium plating onset is revealed. The results lay a good foundation for proposing the online diagnostic method of lithium plating onset based on the in situ impedance.

Accurate detection of dilated cardiomyopathy onset through machine learning predictions from ECG data

Abstract Background Dilated cardiomyopathy (DCM) is a primary heart muscle disease characterised by left-ventricular dilatation and reduced contractility, often associated with arrhythmia and conduction disease. DCM often has a genetic component, putting family members of diagnosed patients at risk for the disease. Stratification tools to provide personalized screening advice are currently lacking. Artificial intelligence (AI) based analysis of the electrocardiogram (ECG) offers a potential solution to this challenge. Purpose Our study aims to develop and validate a novel AI-ECG model capable of prediction and early detection of DCM. Methods An AI-ECG model was developed on a United States (US) based secondary care centre dataset including 337,463 ECGs from 50,932 patients within 4 months of diagnostic echocardiography and 265,576 ECGs from 51,086 patients prior to the first diagnostic echo. The data was split 50/10/40% for training, validation, and testing, with each subset containing 9-12% DCM positive labels. The model architecture was based on a convolutional neural network with residual blocks with a modified final layer for a discrete-time survival model. The AI-ECG score was linked to changes in the median ECG within the test set to provide explainability. External validation was performed on 68,885 diagnostic ECGs (0.08% DCM positive) from the UK biobank (UKB) and a Chinese general hospital dataset with 317,854 diagnostic ECGs (2.2% DCM positive) and 76,936 predictive ECGs (1.7% DCM positive). We performed common and rare variant association studies for the AI ECG scores on quality-controlled UKB data. Results The AI-ECG score for predicted future DCM was associated with diverse ECG features (Fig. 1). In the US cohort test set, the model achieved high diagnostic accuracy with an AUC of 0.9 (95% CI 0.89-0.90) overall, and 0.89 (0.85-0.92) in a subset excluding ischemic or valvular heart disease. External datasets showed AUCs of 0.86 (0.81-0.91) in the UKB and 0.97 (96-0.97) in the Chinese dataset. The predictive performance of the AI-ECG score, adjusted for age and sex, closely approximated echocardiographic traits (LVEF and LVEDD, Fig. 2) with a concordance index (CI) of 0.79 (0.76-0.83) and 0.81 (0.77-0.85), respectively, at 10 years of follow up (FU). Combining AI-ECG scores with echocardiographic traits enhanced the CI to 0.84 (0.8-0.87). Similar findings were observed in the Chinese dataset, with a CI of 0.87 (0.85-0.89) for the AI-ECG score and 0.9 (0.88-0.92) for echocardiographic traits. Genetic analyses in UKB identified associations between the AI-ECG score and rare TTN variants as well as 25 common variant loci. Conclusion Our explainable AI ECG model for future DCM prediction demonstrates robust performance across datasets, providing comparable predictive accuracy to echocardiographic traits up to 10 years pre-diagnosis. Genetic associations support the model's validity and provide new tools for DCM genetic discovery.

Resonance Testing Data Evaluation Approaches for Scaling Onset Detection in Pipelines

Most industries dealing with pipelines face problems resulting from the buildup of deposits therein, such as reduced efficiency, downtime and increased maintenance costs. Although solutions to this issue have been sought for decades, no widely employed technique for monitoring growth of inorganic deposits (or ‘scaling’) in pipelines exists. In this research, a means of detecting the onset of scaling growth, by processing resonance testing data, was sought. For the resonance testing measurements the pipeline segment of interest is equipped with acceleration sensors which record signals generated by impacting the pipeline with a steel tip. The signals are Fourier transformed and analysed in the frequency domain, where a clear shift in frequency peak positions can be observed as the scaling thickness changes. How best to extract quantitative information from the generated frequency data is an open question. In this research, two data analysis approaches for scaling thickness prediction are compared: a supervised (binary classification) machine learning approach as well as a comparison-based change detection approach using cross-correlation. The supervised machine learning approach yields generalizable results for different acceleration sensors and impactor diameters whilst the change detection approach is sensitive from a scaling thickness of 0.5 mm. Whilst this research is specific to the pipe–scaling geometry—and type used in the experiments conducted, resonance testing can be applied to any pipe–scaling combination. The robustness of the data processing approaches presented in this work, when applied to other pipe–scaling materials and geometries, is the next point of research.

Feature selection based on game theory optimization to achieve desired performance metrics in seizure onset detection

Background and Objective: Epilepsy is one of the common diseases of the nervous system, which affects almost 1% of the world’s population. The unpredictable nature of epileptic seizures has caused people suffering from it to be deprived of their normal life, and making a device that can detect a seizure onset can help them. Methods:Game theory is a branch of mathematical science that, after proving its ability in economics, has entered into the field of signal processing and is trying to solve complex problems with a new approach. In this paper, we use game theory and coalitional game concepts to provide a novel way of feature selection in the analysis of Electroencephalography (EEG) signals. We have used a variety of statistical, morphological, and frequency features, and with the proposed feature selection algorithm we have attempted to select the smallest coalition which discriminates between two brain states while appropriately allocating resources to satisfy our desired sensitivity, delay, and false alarm. A modified change point detection algorithm has also been introduced to be used to detect epileptic seizures. Results:The results of the test on the CHB-MIT dataset show that using the proposed algorithm, we can achieve improved performance in terms of sensitivity, delay, and false alarm with a few number of features. Our framework also provides a means to focus on desired criteria for different applications. In scenario which the focus was on sensitivity, the proposed framework achieved 99% average sensitivity, 0.4/h average false alarm, and 2.4 s average delay with only using about 2 features for each patient. Conclusion:By applying the proposed framework on the CHB-MIT dataset we observe that we can achieve better performance compared to recent studies with fewer features in terms of sensitivity, delay, and false alarm metrics. Such an algorithm can be used for the purpose of real-time detection and in wearable devices.

Inducing ipsilateral motor-evoked potentials in the biceps brachii muscle in healthy humans.

To assess reticulospinal tract excitability, high-intensity transcranial magnetic stimulation (TMS) has been used to elicit ipsilateral motor-evoked potentials (iMEPs). However, there is no consensus on robust and valid methods for use in human studies. The present study proposes a standardized method for eliciting and analysing iMEPs in the biceps brachii. Twenty-four healthy young adults participated in this study. Electromyography (EMG) electrodes recorded contralateral MEPs (cMEPs) from the right and iMEPs from the left biceps brachii. A dynamic preacher curl task was used with ~15% of the subject's one-repetition maximum load. The protocol included maximal compound action potential (M-max) determination of the right biceps brachii muscle, TMS hotspot determination, and four sets of five repetitions where 100% stimulator output was delivered at an elbow angle of 110° of flexion. We normalized cMEP amplitude by M-max (% M-max) and iMEP by cMEP amplitude ratio (ICAR). Clear iMEPs above background EMG were observed in 21 subjects (88%, ICAR = .31 ± .19). Good-to-excellent agreement (intraclass correlation coefficient [ICC] = .795-1.000) and low bias (.01-.08 mV and .60-1.11 ms) were demonstrated when comparing two different analysis methods (i.e. fixed time-window vs. manual onset detection) to determine the cMEP and iMEP amplitude and latency, respectively. Most subjects demonstrated clear iMEPs above background EMG triggered at a pre-determined joint angle during a light-load dynamic preacher curl exercise. Similar results were obtained when comparing a single-trial manual identification of iMEP and a semi-automated time-window data analysis approach.

Spatial lung imaging in clinical and translational settings.

For many severe lung diseases, non-invasive biomarkers from imaging could improve early detection of lung injury or disease onset, establish a diagnosis, or help follow-up disease progression and treatment strategies. Imaging of the thorax and lung is challenging due to its size, respiration movement, transferred cardiac pulsation, vast density range and gravitation sensitivity. However, there is extensive ongoing research in this fast-evolving field. Recent improvements in spatial imaging have allowed us to study the three-dimensional structure of the lung, providing both spatial architecture and transcriptomic information at single-cell resolution. This fast progression, however, comes with several challenges, including significant image file storage and network capacity issues, increased costs, data processing and analysis, the role of artificial intelligence and machine learning, and mechanisms to combine several modalities. In this review, we provide an overview of advances and current issues in the field of spatial lung imaging.

Muscle pre-activation prior to landing in athletes with anterior cruciate ligament reconstruction: detection of EMG onset using artificial intelligence

Muscle pre-activation prior to landing in athletes with anterior cruciate ligament reconstruction: detection of EMG onset using artificial intelligence

DeepTool: A deep learning framework for tool wear onset detection and remaining useful life prediction

Milling tool availability and its useful life estimation is essential for optimisation, reliability and cost reduction in milling operations. This work presents DeepTool, a deep learning-based system that predicts the service life of the tool and detects the onset of its wear. DeepTool showcases a comprehensive feature extraction process, and a self-collected dataset of sensor data from milling tests carried out under different cutting settings to extract relevant information from the sensor signals. The main contributions of this study are:•Self-Collected Dataset: Makes use of an extensive, self-collected dataset to record precise sensor signals during milling.•Advanced Predictive Modeling: Employs hybrid autoencoder-LSTM and encoder-decoder LSTM models to estimate tool wear onset and predict its remaining useful life with over 95 % R2 accuracy score.•Comprehensive Feature Extraction: Employs an efficient feature extraction technique from the gathered sensor data, emphasising both time-domain and frequency-domain aspects associated with tool wear.

Comparison of Three Influenza Surveillance Data Sources for Timely Detection of Epidemic Onset - Chengdu City, Sichuan Province and Beijing Municipality, China, 2017-2023.

The syndromic surveillance system, exemplified by the influenza-like illness (ILI) surveillance system, has long been crucial in providing early warnings of influenza epidemics. The analysis revealed that employing reported influenza case data from the nationwide Notifiable Infectious Diseases Reporting Information System (NIDRIS) enhanced the early detection of influenza epidemics, particularly within the context of multiple respiratory pathogens circulating concurrently. The NIDRIS, characterized by its extensive coverage, obligatory reporting, high specificity, and real-time data transmission, offers a valuable tool for the effective early detection of influenza epidemics. Utilizing this system could enhance preparedness and responses to such health crises, potentially mitigating their impact on public health.

Data-Driven Approach to the Development of an Aeroelastic Flutter Precursor

This paper presents a novel algorithm for aeroelastic flutter early detection. Two new features for flutter onset detection are presented. Flutter early warning is accomplished using only measured signals, with essentially no prior knowledge needed about the aircraft or the flutter mechanism involved. The algorithm consists of three stages: 1) extraction of regularity features, 2) calibration by addition of white noise to nominal measurements, and 3) thresholding. Four types of datasets were used: a) synthetic data, b) simulated data generated using aeroelastic response simulations to stochastic gusts, c) measured data from a wind tunnel experiment, and d) flight test data including actual flutter onsets. The algorithm was shown to be able to flag an impending flutter event before critical onset occurs. (The Python code for paper is available at https://github.com/bmeivar/flutter.)

Evaluation of non-occlusive mesenteric ischemia for burn patients – A matched-pair analysis and treatment algorithm

BackgroundBurn injuries may cause gastrointestinal dysfunction leading to intestinal barrier dysfunction, abdominal compartment syndrome, and acute mesenteric ischemia. In the absence of major vascular occlusion, non-occlusive mesenteric ischemia (NOMI) often occurs in critically ill intensive-care burn patients. MethodsA retrospective descriptive analysis of the burn registry of the Department of Plastic, Aesthetic, Hand and Reconstructive Surgery of Hannover Medical School was performed from 1st January 2018 to 1st May 2024. Burn patients with NOMI were matched with burn patients who did not acquire acute mesenteric ischemia based on key variables and shared characteristics. ResultsA total of 20 patients were included in this study. Patients with NOMI showed a statistically significant elevation in serum lactate (p = 0.005) and were most likely to be in a shock state requiring vasopressors (p = 0.047). Overall prognosis was poor for the NOMI cohort, 80 % of whom had a fatal result (p = 0.024). A total of four patients received intra-arterial administration of alprostadil. ConclusionsNOMI represents a potentially fatal condition for the burn patient. The current lack of sensitive biomarkers and accurate diagnostic tools for the early detection of NOMI onset is a major factor behind the overall poor prognosis. We propose the intra-arterial administration of alprostadil as a novel approach to targeted treatment for NOMI.

The assessment method of lip closure ability based on sEMG nonlinear onset detection algorithms.

To overcome the limitations of traditional diagnosis of orbicularis oris muscle function in mouth-breathing patients, this study aims to propose a surface electromyographic (sEMG) based method for reliable and accurate quantitative assessment of lip closure ability. A total of 21 volunteers (16 patients and 5 healthy subjects, aged 8-16) were included in the study. Three nonlinear onset detection algorithms- Teager-Kaiser Energy (TKE) operator, Sample Entropy (SampEn), and Fuzzy Entropy (FuzzyEn)- were compared for their ability to identify lip closure in sEMG signals. Lip Closure EMG Activity Index (LCEAI) was proposed based on the action segments detected by the best performing algorithm for the quantitative assessment of lip closure. The results indicated that FuzzyEn had the highest lip closure identification rate at 93.78 %, the lowest average onset delay of 47.50 ms, the lowest average endpoint delay of 73.10 ms, and the minimal time error of 111.61 ms, exhibiting superior performance. The calculation results of the LCEAI closely corresponded with the actual degree of lip closure in patients. The lip closure ability assessment method proposed in this study can provide a quantitative basis for the diagnosis of mouth breathing.

Evaluation of the performance of meteorological drought indices in Morocco: A case study of different climatic zones

AbstractUnderstanding drought occurrence and evolution is important in minimizing the impacts associated with it. This work assesses the performance of 10 commonly used meteorological indices to measure drought in Morocco. The studied indices are Deciles Index (DI), Percent of Normal Index (PNI), Z‐Score Index (ZSI), China‐Z Index (CZI), Rainfall Anomaly Index (RAI), Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), Palmer Drought Severity Index (PDSI), its self‐calibrated variant (scPDSI) and Palmer Z Index (PZI). Rainfall and temperature gridded data is sourced from PERSIANN‐CDR and ERA5, respectively, for the period 1983–2021. The study area exhibits three main climatic regimes; subhumid, semi‐arid and arid, with a drying and warming climate, as depicted by the rainfall and temperature trends analysis. Results show that most rainfall‐based indices perform relatively poorly in drought monitoring in the study area. DI and PNI appear to be inconsistent and abnormally responsive to rainfall. RAI reports droughts 56.5% more frequently and slightly underestimate drought intensity compared to other indices. Similarly, ZSI and CZI largely underestimate drought intensity. PDSI and scPDSI are computationally demanding, often underestimate drought intensity and overestimate drought duration by at least 115% compared to SPI and SPEI. Conversely, PZI can be used for drought onset detection as it reported droughts early compared to the other indices. SPI and SPEI perform overall better regarding their consistent drought identification and severity assessment. However, SPEI is found to be more suitable than SPI in the arid and semi‐arid regions and performed better considering the warming climate of the country.

Effect of Sampling Rate, Filtering, and Torque Onset Detection on Quadriceps Rate of Torque Development and Torque Steadiness.

Quadriceps rate of torque development (RTD) and torque steadiness are valuable metrics for assessing explosive strength and the ability to control force over a sustained period of time, which can inform clinical assessments of knee function. Despite their widespread use, there is a significant gap in standardized methodology for measuring these metrics, which limits their utility in comparing outcomes across different studies and populations. To address these gaps, we evaluated the influence of sampling rates, signal filtering, and torque onset detection on RTD and torque steadiness. Twenty-seven participants with a history of a primary anterior cruciate ligament reconstruction (N = 27 (11 male/16 female), age = 23 ± 8 years, body mass index = 26 ± 4 kg/m2) and thirty-two control participants (N = 32 (13 male/19 female), age = 23 ± 7 years, body mass index = 23 ± 3 kg/m2) underwent isometric quadriceps strength testing, with data collected at 2222 Hz on an isokinetic dynamometer. The torque-time signal was downsampled to approximately 100 and 1000 Hz and processed using a low-pass, zero-lag Butterworth filter with a range of cutoff frequencies spanning 10-200 Hz. The thresholds used to detect torque onset were defined as 0.1 Nm, 1 Nm, and 5 Nm. RTD between 0 and 100 ms, 0 and 200 ms, and 40-160 ms was computed, as well as absolute and relative torque steadiness. Relative differences were computed by comparing all outcomes to the "gold standard" values computed, with a sampling rate of 2222 Hz, a cutoff frequency in the low-pass filter of 150 Hz, and torque onset of 1 Nm, and compared utilizing linear mixed models. While all combinations of signal collection and processing parameters reached statistical significance (p < 0.05), these differences were consistent between injured and control limbs. Additionally, clinically relevant differences (+/-10%) were primarily observed through torque onset detection methods and primarily affected RTD between 0 and 100 ms. Although measurements of RTD and torque steadiness were generally robust against diverse signal collection and processing parameters, the selection of torque onset should be carefully considered, especially in early RTD assessments that have shorter time epochs.

Detecting Melt Pond Onset on Landfast Arctic Sea Ice Using a Dual C-Band Satellite Approach

The presence of melt ponds on the surface of Arctic Sea ice affects its albedo, thermal properties, and overall melting rate; thus, the detection of melt pond onset is of significant importance for understanding the Arctic’s changing climate. This study investigates the utility of a novel method for detecting the onset of melt ponds on sea ice using a satellite-based, dual-sensor C-band approach, whereby Sentinel-1 provides horizontally polarized (HH) data and Advanced SCATterometer (ASCAT) provides vertically polarized (VV) data. The co-polarized ratio (VV/HH) is used to detect the presence of melt ponds on landfast sea ice in the Canadian Arctic Archipelago in 2017 and 2018. ERA-5 air temperature and wind speed re-analysis datasets are used to establish the VV/HH threshold for pond onset detection, which have been further validated by Landsat-8 reflectance. The co-polarized ratio threshold of three standard deviations from the late winter season (April) mean co-pol ratio values are used for assessing pond onset detection associated with the air temperature and wind speed data, along with visual observations from Sentinel-1 and cloud-free Sentinel-2 imagery. In 2017, the pond onset detection rates were 70.59% for FYI and 92.3% for MYI. Results suggest that this method, because of its dual-platform application, has potential for providing large-area coverage estimation of the timing of sea ice melt pond onset using different earth observation satellites.

Developing a Multivariate Agro‐Meteorological Index to Improve Capturing Onset and Persistence of Droughts Utilizing Vapor Pressure Deficit and Soil Moisture

AbstractDrought is associated with adverse environmental and societal impacts across various regions. Therefore, drought monitoring based on a single variable may lead to unreliable information, especially about the onset and persistence of drought. Previous studies show vapor pressure deficit (VPD) data can detect drought onset earlier than other drought indicators such as precipitation. On the other hand, soil moisture (SM) is a robust indicator for assessing drought persistence. This study introduces a nonparametric multivariate drought index Vapor Pressure Deficit Soil moisture standardized Drought Index (VPDSDI) which is developed by combining VPD with SM information. The performance of the multivariate index in terms of drought onset detection is compared with the Standardized Precipitation Evapotranspiration Index (SPEI) and the Standardized Precipitation Index (SPI) for six major drought events across the United States including three rapidly developing drought events (this term refers to flash droughts that develop on monthly scales) and three conventional drought events. Additionally, the performance of the proposed index in detecting drought persistence is compared with the Standardized Soil moisture Index (SSI), which is an agricultural drought index. Results indicate the multivariate index detects drought onset always earlier than SPI for conventional events, but VPDSDI detects drought onset earlier than or about the same time as SPEI for rapidly developing droughts. In terms of persistence, VPDSDI detects persistence almost identical to SSI for both rapidly developing and conventional drought events. The results also show that combining VPD with SM reduces the high variability of VPD and produces a smoother index which improves the onset and persistence detection of drought events leveraging VPD and SM information.

Reducing False Alarms in Wearable Seizure Detection With EEGformer: A Compact Transformer Model for MCUs.

The long-term, continuous analysis of electroencephalography (EEG) signals on wearable devices to automatically detect seizures in epileptic patients is a high-potential application field for deep neural networks, and specifically for transformers, which are highly suited for end-to-end time series processing without handcrafted feature extraction. In this work, we propose a small-scale transformer detector, the EEGformer, compatible with unobtrusive acquisition setups that use only the temporal channels. EEGformer is the result of a hardware-oriented design exploration, aiming for efficient execution on tiny low-power micro-controller units (MCUs) and low latency and false alarm rate to increase patient and caregiver acceptance.Tests conducted on the CHB-MIT dataset show a 20% reduction of the onset detection latency with respect to the state-of-the-art model for temporal acquisition, with a competitive 73% seizure detection probability and 0.15 false-positive-per-hour (FP/h). Further investigations on a novel and challenging scalp EEG dataset result in the successful detection of 88% of the annotated seizure events, with 0.45 FP/h.We evaluate the deployment of the EEGformer on three commercial low-power computing platforms: the single-core Apollo4 MCU and the GAP8 and GAP9 parallel MCUs. The most efficient implementation (on GAP9) results in as low as 13.7 ms and 0.31 mJ per inference, demonstrating the feasibility of deploying the EEGformer on wearable seizure detection systems with reduced channel count and multi-day battery duration.

Early adverse physiological event detection using commercial wearables: challenges and opportunities

Data from commercial off-the-shelf (COTS) wearables leveraged with machine learning algorithms provide an unprecedented potential for the early detection of adverse physiological events. However, several challenges inhibit this potential, including (1) heterogeneity among and within participants that make scaling detection algorithms to a general population less precise, (2) confounders that lead to incorrect assumptions regarding a participant’s healthy state, (3) noise in the data at the sensor level that limits the sensitivity of detection algorithms, and (4) imprecision in self-reported labels that misrepresent the true data values associated with a given physiological event. The goal of this study was two-fold: (1) to characterize the performance of such algorithms in the presence of these challenges and provide insights to researchers on limitations and opportunities, and (2) to subsequently devise algorithms to address each challenge and offer insights on future opportunities for advancement. Our proposed algorithms include techniques that build on determining suitable baselines for each participant to capture important physiological changes and label correction techniques as it pertains to participant-reported identifiers. Our work is validated on potentially one of the largest datasets available, obtained with 8000+ participants and 1.3+ million hours of wearable data captured from Oura smart rings. Leveraging this extensive dataset, we achieve pre-symptomatic detection of COVID-19 with a performance receiver operator characteristic (ROC) area under the curve (AUC) of 0.725 without correction techniques, 0.739 with baseline correction, 0.740 with baseline correction and label correction on the training set, and 0.777 with baseline correction and label correction on both the training and the test set. Using the same respective paradigms, we achieve ROC AUCs of 0.919, 0.938, 0.943 and 0.994 for the detection of self-reported fever, and 0.574, 0.611, 0.601, and 0.635 for detection of self-reported shortness of breath. These techniques offer improvements across almost all metrics and events, including PR AUC, sensitivity at 75% specificity, and precision at 75% recall. The ring allows continuous monitoring for detection of event onset, and we further demonstrate an improvement in the early detection of COVID-19 from an average of 3.5 days to an average of 4.1 days before a reported positive test result.

Physiologic validation of the Compensatory Reserve Metric obtained from pulse oximetry: A step toward advanced medical monitoring on the battlefield.

The Compensatory Reserve Metric (CRM) provides a time sensitive indicator of hemodynamic decompensation. However, its in-field utility is limited because of the size and cost-intensive nature of standard vital sign monitors or photoplethysmographic volume-clamp (PPG VC ) devices used to measure arterial waveforms. In this regard, photoplethysmographic measurements obtained from pulse oximetry may serve as a useful, portable alternative. This study aimed to validate CRM values obtained using pulse oximeter (PPG PO ). Forty-nine healthy adults (25 females) underwent a graded lower body negative pressure (LBNP) protocol to simulate hemorrhage. Arterial waveforms were sampled using PPG PO and PPG VC . The CRM was calculated using a one-dimensional convolutional neural network. Cardiac output and stroke volume were measured using PPG VC . A brachial artery catheter was used to measure intra-arterial pressure. A three-lead electrocardiogram was used to measure heart rate. Fixed-effect linear mixed models with repeated measures were used to examine the association between CRM values and physiologic variables. Log-rank analyses were used to examine differences in shock determination during LBNP between monitored hemodynamic parameters. The median LBNP stage reached was 70 mm Hg (range, 45-100 mm Hg). Relative to baseline, at tolerance, there was a 47% ± 12% reduction in stroke volume, 64% ± 27% increase in heart rate, and 21% ± 7% reduction in systolic blood pressure ( p < 0.001 for all). Compensatory Reserve Metric values obtained with both PPG PO and PPG VC were associated with changes in heart rate ( p < 0.001), stroke volume ( p < 0.001), and pulse pressure ( p < 0.001). Furthermore, they provided an earlier detection of hemodynamic shock relative to the traditional metrics of shock index ( p < 0.001 for both), systolic blood pressure ( p < 0.001 for both), and heart rate ( p = 0.001 for both). The CRM obtained from PPG PO provides a valid, time-sensitized prediction of hemodynamic decompensation, opening the door to provide military medical personnel noninvasive in-field advanced capability for early detection of hemorrhage and imminent onset of shock. Diagnostic Tests or Criteria; Level III.