Application of a Bayesian Statistical Framework for Planetary Protection as a Means of Verifying Low-Biomass, Zero-Inflated Test Data from Spacecraft.

Planetary Protection is applicable for missions to biologically sensitive targets of interest in the solar system. For robotic missions landing on the Martian surface, Earth-based biological contamination must be reduced, controlled, and monitored to adhere to forward planetary protection requirements. To address the overall biological load limit and microbial density requirements per spacecraft each component is tracked based on its manufacturing pedigree and/or directly assessed using a direct sampling technique with either a swab or wipe. The tracking and reporting of requirements compliance has varied from mission to mission and reporting of numbers has consistently leaned towards the conservative worst-case scenario. With an increase in the number of missions and mission complexities, the need to establish a technically sound, statistical, and biological solution that provides a single point solution which addresses the distribution of spacecraft contamination becomes critical. Select components of the InSight mission, launched in 2018, have been used as a test case to evaluate the efficacy of applying Bayesian statistics to planetary protection data sets. Eight representative components covering the various bounding cases of high and low surface area, biological count, and sampling devices were analyzed as well as an assembly level case to evaluate the rollup of directly sampled and manufacturing pedigree components. A Bayesian approach was developed leveraging different priors from the zero-inflated data sets and compared to the heritage and existing NASA bioburden assessment approaches. In addition, several non-informative priors were evaluated for use in performing bioburden calculations. The results have demonstrated a viable framework to enable a Bayesian statistical approach to be further developed and utilized for planetary protection requirements assessment.

Identifications and classifications of human locomotion using Rayleigh-enhanced distributed fiber acoustic sensors with deep neural networks.

This paper reports on the use of machine learning to delineate data harnessed by fiber-optic distributed acoustic sensors (DAS) using fiber with enhanced Rayleigh backscattering to recognize vibration events induced by human locomotion. The DAS used in this work is based on homodyne phase-sensitive optical time-domain reflectometry (φ-OTDR). The signal-to-noise ratio (SNR) of the DAS was enhanced using femtosecond laser-induced artificial Rayleigh scattering centers in single-mode fiber cores. Both supervised and unsupervised machine-learning algorithms were explored to identify people and specific events that produce acoustic signals. Using convolutional deep neural networks, the supervised machine learning scheme achieved over 76.25% accuracy in recognizing human identities. Conversely, the unsupervised machine learning scheme achieved over 77.65% accuracy in recognizing events and human identities through acoustic signals. Through integrated efforts on both sensor device innovation and machine learning data analytics, this paper shows that the DAS technique can be an effective security technology to detect and to identify highly similar acoustic events with high spatial resolution and high accuracies.

Distributed fiber sensor and machine learning data analytics for pipeline protection against extrinsic intrusions and intrinsic corrosions.

This paper presents an integrated technical framework to protect pipelines against both malicious intrusions and piping degradation using a distributed fiber sensing technology and artificial intelligence. A distributed acoustic sensing (DAS) system based on phase-sensitive optical time-domain reflectometry (φ-OTDR) was used to detect acoustic wave propagation and scattering along pipeline structures consisting of straight piping and sharp bend elbow. Signal to noise ratio of the DAS system was enhanced by femtosecond induced artificial Rayleigh scattering centers. Data harnessed by the DAS system were analyzed by neural network-based machine learning algorithms. The system identified with over 85% accuracy in various external impact events, and over 94% accuracy for defect identification through supervised learning and 71% accuracy through unsupervised learning.

Estimation of metabolic energy expenditure from core temperature using a human thermoregulatory model.

Human metabolic energy expenditure is critical to many scientific disciplines but can only be measured using expensive and/or restrictive equipment. The aim of this work is to determine whether the SCENARIO thermoregulatory model can be adapted to estimate metabolic rate (M) from core body temperature (TC). To validate this method of M estimation, data were collected from fifteen test volunteers (age = 23 ± 3yr, height = 1.73 ± 0.07m, mass = 68.6 ± 8.7kg, body fat = 16.7 ± 7.3%; mean ± SD) who wore long sleeved nylon jackets and pants (Itot,clo = 1.22, Im = 0.41) during treadmill exercise tasks (32 trials; 7.8 ± 0.5km in 1h; air temp. = 22°C, 50% RH, wind speed = 0.35ms-1). Core body temperatures were recorded by ingested thermometer pill and M data were measured via whole room indirect calorimetry. Metabolic rate was estimated for 5min epochs in a two-step process. First, for a given epoch, a range of M values were input to the SCENARIO model and a corresponding range of TC values were output. Second, the output TC range value with the lowest absolute error relative to the observed TC for the given epoch was identified and its corresponding M range input was selected as the estimated M for that epoch. This process was then repeated for each subsequent remaining epoch. Root mean square error (RMSE), mean absolute error (MAE), and bias between observed and estimated M were 186W, 130 ± 174W, and 33 ± 183W, respectively. The RMSE for total energy expenditure by exercise period was 0.30 MJ. These results indicate that the SCENARIO model is useful for estimating M from TC when measurement is otherwise impractical.

Measuring the short-term substrate utilization response to high-carbohydrate and high-fat meals in the whole-body indirect calorimeter.

The paper demonstrates that minute-to-minute metabolic response to meals with different macronutrient content can be measured and discerned in the whole-body indirect calorimeter. The ability to discriminate between high-carbohydrate and high-fat meals is achieved by applying a modified regularization technique with additional constraints imposed on oxygen consumption rate. These additional constraints reduce the differences in accuracy between the oxygen and carbon dioxide analyzers. The modified technique was applied to 63 calorimeter sessions that were each 24 h long. The data were collected from 16 healthy volunteers (eight males, eight females, aged 22-35 years). Each volunteer performed four 24-h long calorimeter sessions. At each session, they received one of four treatment combinations involving exercise (high or low intensity) and diet (a high-fat or high-carbohydrate shake for lunch). One volunteer did not complete all four assignments, which brought the total number of sessions to 63 instead of 64. During the 24-h stay in the calorimeter, subjects wore a continuous glucose monitoring system, which was used as a benchmark for subject's postprandial glycemic response. The minute-by-minute respiratory exchange ratio (RER) data showed excellent agreement with concurrent subcutaneous glucose concentrations in postprandial state. The averaged minute-to-minute RER response to the high-carbohydrate shake was significantly different from the response to high-fat shake. Also, postprandial RER slopes were significantly different for two dietary treatments. The results show that whole-body respiration calorimeters can be utilized as tools to study short-term kinetics of substrate oxidation in humans.

A single bout of resistance exercise does not promote excess postexercise energy expenditure in untrained young men with a family history of diabetes.

BACKGROUND: People with a family history of type 2 diabetes have lower energy expenditure (EE) and more obesity than those having no such family history. Resistance exercise (RE) may induce excess postexercise energy expenditure (EPEE) and reduce long-term risk for obesity in this susceptible group. PURPOSE: To determine the effect of RE on EPEE for 15 hr after a single exercise bout in healthy, untrained young men having a family history of type 2 diabetes. DESIGN: Seven untrained men (23 ± 1.2 years, BMI 24 ± 1.1) completed a 48-hr protocol in a whole room calorimeter. The first day served as a control day, with a moderate 40-min RE bout occurring on the second day. Differences in postexercise EE were compared with matched periods from the control day for cumulative 15-min intervals (up to 150 min) and 15 hr after the RE bout was completed. RESULTS: The most robust difference in EPEE between the experimental and control days was observed in the first 15-min postexercise period (M = 1.4Kcal/min; SD = 0.7; p < .05). No statistically significant differences in EPEE were noted beyond 90-min of continuous measurement. CONCLUSIONS: Young people with a family history of type 2 diabetes may not show EPEE after a single RE bout when observed for 15 hr after RE and long-term resistance training may be required to promote EPEE.

Three 15-min bouts of moderate postmeal walking significantly improves 24-h glycemic control in older people at risk for impaired glucose tolerance.

OBJECTIVE: The purpose of this study was to compare the effectiveness of three 15-min bouts of postmeal walking with 45 min of sustained walking on 24-h glycemic control in older persons at risk for glucose intolerance. RESEARCH DESIGN AND METHODS: Inactive older (≥60 years of age) participants (N=10) were recruited from the community and were nonsmoking, with a BMI<35 kg/m2 and a fasting blood glucose concentration between 105 and 125 mg dL(-1). Participants completed three randomly ordered exercise protocols spaced 4 weeks apart. Each protocol comprised a 48-h stay in a whole-room calorimeter, with the first day serving as the control day. On the second day, participants engaged in either 1) postmeal walking for 15 min or 45 min of sustained walking performed at 2) 10:30 a.m. or 3) 4:30 p.m. All walking was on a treadmill at an absolute intensity of 3 METs. Interstitial glucose concentrations were determined over 48 h with a continuous glucose monitor. Substrate utilization was measured continuously by respiratory exchange (VCO2/VO2). RESULTS: Both sustained morning walking (127±23 vs. 118±14 mg dL(-1)) and postmeal walking (129±24 vs. 116±13 mg dL(-1)) significantly improved 24-h glycemic control relative to the control day (P<0.05). Moreover, postmeal walking was significantly (P<0.01) more effective than 45 min of sustained morning or afternoon walking in lowering 3-h postdinner glucose between the control and experimental day. CONCLUSIONS: Short, intermittent bouts of postmeal walking appear to be an effective way to control postprandial hyperglycemia in older people.

On the accuracy of instantaneous gas exchange rates, energy expenditure and respiratory quotient calculations obtained from indirect whole room calorimetry.

This paper analyzes the accuracy of metabolic rate calculations performed in the whole room indirect calorimeter using the molar balance equations. The equations are treated from the point of view of cause-effect relationship where the gaseous exchange rates representing the unknown causes need to be inferred from a known, noisy effect-gaseous concentrations. Two methods of such inference are analyzed. The first method is based on the previously published regularized deconvolution of the molar balance equation and the second one, proposed in this paper, relies on regularized differentiation of gaseous concentrations. It is found that both methods produce similar results for the absolute values of metabolic variables and their accuracy. The uncertainty for O2 consumption rate is found to be 7% and for CO2 production--3.2%. The uncertainties in gaseous exchange rates do not depend on the absolute values of O2 consumption and CO2 production. In contrast, the absolute uncertainty in respiratory quotient is a function of the gaseous exchange rates and varies from 9.4% during the night to 2.3% during moderate exercise. The uncertainty in energy expenditure was found to be 5.9% and independent of the level of gaseous exchange. For both methods, closed form analytical formulas for confidence intervals are provided allowing quantification of uncertainty for four major metabolic variables in real world studies.

Are standard diagnostic test characteristics sufficient for the assessment of continual patient monitoring?

BACKGROUND: For diagnostic processes involving continual measurements from a single patient, conventional test characteristics, such as sensitivity and specificity, do not consider decision consistency, which might be a distinct, clinically relevant test characteristic. OBJECTIVE: The authors investigated the performance of a decision-support classifier for the diagnosis of traumatic injury with blood loss, implemented with three different data-processing methods. For each method, they computed standard diagnostic test characteristics and novel metrics related to decision consistency and latency. SETTING: Prehospital air ambulance transport. PATIENTS: A total of 557 trauma patients. DESIGN: Continually monitored vital-sign data from 279 patients (50%) were randomly selected for classifier development, and the remaining were used for testing. Three data-processing methods were evaluated over 16 min of patient monitoring: a 2-min moving window, time averaging, and postprocessing with the sequential probability ratio test (SPRT). MEASUREMENTS: Sensitivity and specificity were computed. Consistency was quantified through cumulative counts of decision changes over time and the fraction of patients affected by false alarms. Latency was evaluated by the fraction of patients without a decision. RESULTS: All 3 methods showed very similar final sensitivities and specificities. Yet, there were significant differences in terms of the fraction of patients affected by false alarms, decision changes through time, and latency. For instance, use of the SPRT led to a 75% reduction in the number of decision changes and a 36% reduction in the number of patients affected by false alarms, at the expense of 3% unresolved final decisions. CONCLUSION: The proposed metrics of decision consistency and decision latency provided additional information beyond what could be obtained from test sensitivity and specificity and are likely to be clinically relevant in some applications involving temporal decision making.

Response to acute psychophysical stress and 24-hour glycemic control in healthy older people.

We examined the relation between stress reactivity and 24 h glycemic control in 17 inactive, healthy older people (≥60 years) under both a novel psychophysical stress and a seated control condition. Plasma cortisol was measured over the course of the stress and recovery periods. Glycemic control was determined over the subsequent 3 h from an oral glucose tolerance test (OGTT) and over 24 h via continuous glucose monitoring (CGM). We observed significant (P < 0.05) elevations in perceived stress, cardiovascular activity, and peak cortisol response at 30 min (10.6 ± 3.1 versus 8.6 ± 2.6 µg·dL(-1), resp.) during the stress compared with the control condition; however, 3 h OGTT glucose and insulin responses were similar between conditions. The CGM data suggested a 30-40 min postchallenge delay in peak glucose response and attenuated glucose clearance over the 6 h following the stress condition, but these alterations were not statistically significant. Healthy older people may demonstrate minimal disruption in metabolic resiliency following everyday psychological stress.

A robust method to estimate instantaneous heart rate from noisy electrocardiogram waveforms.

We propose a new algorithm for real-time estimation of instantaneous heart rate (HR) from noise-laden electrocardiogram (ECG) waveforms typical of unstructured, ambulatory field environments. The estimation of HR from ECG waveforms is an indirect measurement problem that requires differencing, which invariably amplifies high-frequency noise. We circumvented noise amplification by considering the estimation of HR as the solution of a weighted regularized least squares problem, which, in addition, directly provided analytically based confidence intervals (CIs) for the estimated HRs. To evaluate the performance of the proposed algorithm, we applied it to simulated data and to noise-laden ECG records that were collected during helicopter transport of trauma-injured patients to a trauma center. We compared the proposed algorithm with HR estimates produced by a widely used vital-sign travel monitor and a standard HR estimation technique, followed by postprocessing with Kalman filtering or spline smoothing. The simulation results indicated that our algorithm consistently produced more accurate HR estimates, with estimation errors as much as 67% smaller than those attained by the postprocessing methods, while the results with the field-collected data showed that the proposed algorithm produced much smoother and reliable HR estimates than those obtained by the vital-sign monitor. Moreover, the obtained CIs reflected the amount of noise in the ECG recording and could be used to statistically quantify uncertainties in the HR estimates. We conclude that the proposed method is robust to different types of noise and is particularly suitable for use in ambulatory environments where data quality is notoriously poor.

Exploration of prehospital vital sign trends for the prediction of trauma outcomes.

OBJECTIVES: We explored whether there are diagnostically useful temporal trends in prehospital vital signs of trauma patients. METHODS: Vital signs were monitored during transport to a level I trauma center and electronically archived. Retrospectively, we identified reliable vital signs recorded from the 0- to 7-minute interval and from the 14- to 21-minute interval during transport, and, for each subject, we computed the temporal differences between the two intervals' vital signs, the intrasubject 95% data ranges, the values during the initial 2 minutes, and the 21-minute overall means. We tested for differences between subjects with major hemorrhage versus control subjects, and computed receiver-operating characteristic (ROC) curves. We conducted sensitivity analyses, exploring alternative clinical outcomes, temporal windows, and methods of identifying reliable data. RESULTS: Comparing major hemorrhage cases versus controls, there were no discriminatory differences in temporal vital sign trends. Hemorrhage cases had significantly wider intrasubject data ranges for systolic blood pressure (SBP), respiratory rate (RR), and shock index (SI) versus controls. All results were consistent in several sensitivity analyses. CONCLUSIONS: Our findings add to a growing body of evidence that prehospital vital sign trends over 21 minutes or less are unlikely to be diagnostically useful because of substantial nondirectional fluctuations in vital signs that would obscure any subtle, progressive temporal trends. SBP, RR, and SI values were significantly different for high-acuity patients, and had more variability. Taken together, these findings suggest that higher-acuity patients experience episodes of instability rather than gradual, steady decline. Measures that account for data variability, such as taking the average of multiple measurements, may improve the diagnostic utility of prehospital vital signs.

The importance of different frequency bands in predicting subcutaneous glucose concentration in type 1 diabetic patients.

We investigated the relative importance and predictive power of different frequency bands of subcutaneous glucose signals for the short-term (0-50 min) forecasting of glucose concentrations in type 1 diabetic patients with data-driven autoregressive (AR) models. The study data consisted of minute-by-minute glucose signals collected from nine deidentified patients over a five-day period using continuous glucose monitoring devices. AR models were developed using single and pairwise combinations of frequency bands of the glucose signal and compared with a reference model including all bands. The results suggest that: for open-loop applications, there is no need to explicitly represent exogenous inputs, such as meals and insulin intake, in AR models; models based on a single-frequency band, with periods between 60-120 min and 150-500 min, yield good predictive power (error <3 mg/dL) for prediction horizons of up to 25 min; models based on pairs of bands produce predictions that are indistinguishable from those of the reference model as long as the 60-120 min period band is included; and AR models can be developed on signals of short length (approximately 300 min), i.e., ignoring long circadian rhythms, without any detriment in prediction accuracy. Together, these findings provide insights into efficient development of more effective and parsimonious data-driven models for short-term prediction of glucose concentrations in diabetic patients.

A real-time algorithm for predicting core temperature in humans.

In this paper, we present a real-time implementation of a previously developed offline algorithm for predicting core temperature in humans. The real-time algorithm uses a zero-phase Butterworth digital filter to smooth the data and an autoregressive (AR) model to predict core temperature. The performance of the algorithm is assessed in terms of its prediction accuracy, quantified by the root mean squared error (RMSE), and in terms of prediction uncertainty, quantified by statistically based prediction intervals (PIs). To evaluate the performance of the algorithm, we simulated real-time implementation using core-temperature data collected during two different field studies, involving ten different individuals. One of the studies includes a case of heat illness suffered by one of the participants. The results indicate that although the real-time predictions yielded RMSEs that are larger than those of the offline algorithm, the real-time algorithm does produce sufficiently accurate predictions for practically meaningful prediction horizons (approximately 20 min). The algorithm reached alert (39 degrees C) and alarm (39.5 degrees C) thresholds for the heat-ill individual but did not even attain the alert threshold for the other individuals, demonstrating the algorithm's good sensitivity and specificity. The PIs reflected, in an intuitively expected manner, the uncertainty associated with real-time forecast as a function of prediction horizon and core-temperature variability. The results also corroborate the feasibility of "universal" AR models, where an offline-developed model based on one individual's data could be used to predict any other individual in real time. We conclude that the real-time implementation of the algorithm confirms the attributes observed in the offline version and, hence, could be considered as a warning tool for impending heat illnesses.

Is respiration-induced variation in the photoplethysmogram associated with major hypovolemia in patients with acute traumatic injuries?

It has been widely accepted that metrics related to respiration-induced waveform variation (RIWV) of the photoplethysmogram (PPG) have been associated with hypovolemia in mechanically ventilated patients and in controlled laboratory environments. In this retrospective study, we investigated if PPG RIWV metrics have diagnostic value for patients with acute hemorrhagic hypovolemia in the prehospital environment. Photoplethysmogram waveforms and basic vital signs were recorded in trauma patients during prehospital transport. Retrospectively, we used automated algorithms to select patient records with all five basic vital signs and 45 s or longer continuous, clean PPG segments. From these segments, we identified the onset and peak of individual heartbeats and computed waveform variations in the beats' peaks and amplitudes: (1) as the range between the maximum and the minimum (max-min) values and (2) as their interquartile range (IQR). We evaluated their receiver operating characteristic (ROC) curves for major hemorrhage. Separately, we tested whether RIWV metrics have potential independent information beyond basic vital signs by applying multivariate regression. In 344 patients, RIWV max-min yielded areas under the ROC curves (AUCs) not significantly better than a random AUC of 0.50. Respiration-induced waveform variation computed as IQR yielded ROC AUCs of 0.65 (95% confidence interval, 0.54-0.76) and of 0.64 (0.51-0.75), for peak and amplitude measures, respectively. The IQR metrics added independent information to basic vital signs (P < 0.05), but only moderately improved the overall AUC. Photoplethysmogram RIWV measured as IQR is preferable over max-min, and using PPG RIWV may enhance physiologic monitoring of spontaneously breathing patients outside strictly controlled laboratory environments.

Universal glucose models for predicting subcutaneous glucose concentration in humans.

This paper tests the hypothesis that a "universal," data-driven model can be developed based on glucose data from one diabetic subject, and subsequently applied to predict subcutaneous glucose concentrations of other subjects, even of those with different types of diabetes. We employed three separate studies, each utilizing a different continuous glucose monitoring (CGM) device, to verify the model's universality. Two out of the three studies involved subjects with type 1 diabetes and the other one with type 2 diabetes. We first filtered the subcutaneous glucose concentration data by imposing constraints on their rate of change. Then, using the filtered data, we developed data-driven autoregressive models of order 30, and used them to make short-term, 30-min-ahead glucose-concentration predictions. We used same-subject model predictions as a reference for comparisons against cross-subject and cross-study model predictions, which were evaluated using the root-mean-squared error (RMSE) and Clarke error grid analysis (EGA). We found that, for each studied subject, the average cross-subject and cross-study RMSEs of the predictions were small and indistinguishable from those obtained with the same-subject models. These observations were corroborated by EGA, where better than 99.0% of the paired sensor-predicted glucose concentrations lay in the clinically acceptable zone A. In addition, the predictive capability of the models was found not to be affected by diabetes type, subject age, CGM device, and interindividual differences. We conclude that it is feasible to develop universal glucose models that allow for clinical use of predictive algorithms and CGM devices for proactive therapy of diabetic patients.

An improved methodology for individualized performance prediction of sleep-deprived individuals with the two-process model.

We present a method based on the two-process model of sleep regulation for developing individualized biomathematical models that predict performance impairment for individuals subjected to total sleep loss. This new method advances our previous work in two important ways. First, it enables model customization to start as soon as the first performance measurement from an individual becomes available. This was achieved by optimally combining the performance information obtained from the individual's performance measurements with a priori performance information using a Bayesian framework, while retaining the strategy of transforming the nonlinear optimization problem of finding the optimal estimates of the two-process model parameters into a series of linear optimization problems. Second, by taking advantage of the linear representation of the two-process model, this new method enables the analytical computation of statistically based measures of reliability for the model predictions in the form of prediction intervals. Two distinct data sets were used to evaluate the proposed method. Results using simulated data with superimposed white Gaussian noise showed that the new method yielded 50% to 90% improvement in parameter-estimate accuracy over the previous method. Moreover, the accuracy of the analytically computed prediction intervals was validated through Monte Carlo simulations. Results for subjects representing three sleep-loss phenotypes who participated in a laboratory study (82 h of total sleep loss) indicated that the proposed method yielded individualized predictions that were up to 43% more accurate than group-average prediction models and, on average, 10% more accurate than individualized predictions based on our previous method.

Exploration of prehospital vital sign trends for the prediction of trauma outcomes.

OBJECTIVES: We explored whether there are diagnostically useful temporal trends in prehospital vital signs of trauma patients. METHODS: Vital signs were monitored during transport to a level I trauma center and electronically archived. Retrospectively, we identified reliable vital signs recorded from the 0- to 7-minute interval and from the 14- to 21-minute interval during transport, and, for each subject, we computed the temporal differences between the two intervals' vital signs, the intrasubject 95% data ranges, the values during the initial 2 minutes, and the 21-minute overall means. We tested for differences between subjects with major hemorrhage versus control subjects, and computed receiver-operating characteristic (ROC) curves. We conducted sensitivity analyses, exploring alternative clinical outcomes, temporal windows, and methods of identifying reliable data. RESULTS: Comparing major hemorrhage cases versus controls, there were no discriminatory differences in temporal vital sign trends. Hemorrhage cases had significantly wider intrasubject data ranges for systolic blood pressure (SBP), respiratory rate (RR), and shock index (SI) versus controls. All results were consistent in several sensitivity analyses. CONCLUSIONS: Our findings add to a growing body of evidence that prehospital vital sign trends over 21 minutes or less are unlikely to be diagnostically useful because of substantial nondirectional fluctuations in vital signs that would obscure any subtle, progressive temporal trends. SBP, RR, and SI values were significantly different for high-acuity patients, and had more variability. Taken together, these findings suggest that higher-acuity patients experience episodes of instability rather than gradual, steady decline. Measures that account for data variability, such as taking the average of multiple measurements, may improve the diagnostic utility of prehospital vital signs.

Predicting subcutaneous glucose concentration in humans: data-driven glucose modeling.

The combination of predictive data-driven models with frequent glucose measurements may provide for an early warning of impending glucose excursions and proactive regulatory interventions for diabetes patients. However, from a modeling perspective, before the benefits of such a strategy can be attained, we must first be able to quantitatively characterize the behavior of the model coefficients as well as the model predictions as a function of prediction horizon. We need to determine if the model coefficients reflect viable physiologic dependencies of the individual glycemic measurements and whether the model is stable with respect to small changes in noise levels, leading to accurate near-future predictions with negligible time lag. We assessed the behavior of linear autoregressive data-driven models developed under three possible modeling scenarios, using continuous glucose measurements of nine subjects collected on a minute-by-minute basis for approximately 5 days. Simulation results indicated that stable and accurate models for near-future glycemic predictions (< 60 min) with clinically acceptable time lags are attained only when the raw glucose measurements are smoothed and the model coefficients are regularized. This study provides a starting point for further needed investigations before real-time deployment can be considered.