Random Measurement Error Research Articles

Test–retest Reliability and Responsiveness of the Machine Learning-based Short-form of the Berg Balance Scale in Persons with Stroke

ObjectiveTo examine the test–retest reliability, responsiveness, and clinical utility of the machine learning-based short-form of the Berg Balance Scale (BBS-ML) in persons with stroke. DesignRepeated measures design. SettingA department of rehabilitation in a medical center. ParticipantsThis study recruited two groups: 50 persons who were more than 6 months post-stroke to examine the test–retest reliability, and 52 persons who were within 3 months post-stroke to examine the responsiveness. Test–retest reliability was investigated by administering assessments twice at a 2-week interval. Responsiveness was investigated by gathering data at admission and discharge from hospital. InterventionsNot applicable. Main outcome measureBBS-ML. ResultsThe BBS-ML exhibited excellent test–retest reliability (intraclass correlation coefficient = 0.99), acceptable minimal random measurement error (minimal detectable change % = 13.6%), and good responsiveness (Kazis' effect size and standardized response mean values ≥ 1.34). On average, the participants completed the BBS-ML in around 6 minutes per administration. ConclusionsOur findings indicate that the BBS-ML appears an efficient measure with excellent test–retest reliability and responsiveness. Moreover, the BBS-ML may be utilized as a substitute for the original BBS to monitor the progress of balance function in persons with stroke.

The Effect of Measurement Error on Hypothesis Testing in Small Sample Structural Equation Modeling: A Comparison of Various Estimation Approaches

Researchers seeking valid statistical inference in the presence of measurement error often apply approaches that ignore measurement error. This may result in biased estimates, inflated type I error rates, diminished power, and therefore, increases the risk of drawing erroneous conclusions. However, current advice on accounting for random measurement error is limited to large samples and traditional linear models. This article aims to address this gap for small samples and recent estimation approaches in structural equation modeling. Our results show substantial type I error rate inflation for approaches that ignore measurement error when the model contains correlated latent predictors.

In silico data-based comparison of the accuracy and error source of various methods for noninvasively estimating central aortic blood pressure

Background and objectivesThe higher clinical significance of central aortic blood pressure (CABP) compared to peripheral blood pressures has been extensively demonstrated. Accordingly, many methods for noninvasively estimating CABP have been proposed. However, there still lacks a systematic comparison of existing methods, especially in terms of how they differ in the ability to tolerate individual differences or measurement errors. The present study was designed to address this gap. MethodsA large-scale ‘virtual subject’ dataset (n = 600) was created using a computational model of the cardiovascular system, and applied to examine several classical CABP estimation methods, including the direct method, generalized transfer function (GTF) method, n-point moving average (NPMA) method, second systolic pressure of periphery (SBP2) method, physical model-based wave analysis (MBWA) method, and suprasystolic cuff-based waveform reconstruction (SCWR) method. The errors of CABP estimation were analyzed and compared among methods with respect to the magnitude/distribution, correlations with physiological/hemodynamic factors, and sensitivities to noninvasive measurement errors. ResultsThe errors of CABP estimation exhibited evident inter-method differences in terms of the mean and standard deviation (SD). Relatively, the estimation errors of the methods adopting pre-trained algorithms (i.e., the GTF and SCWR methods) were overall smaller and less sensitive to variations in physiological/hemodynamic conditions and random errors in noninvasive measurement of brachial arterial blood pressure (used for calibrating peripheral pulse wave). The performances of all the methods worsened following the introduction of random errors to peripheral pulse wave (used for deriving CABP), as characterized by the enlarged SD and/or increased mean of the estimation errors. Notably, the GTF and SCWR methods did not exhibit a better capability of tolerating pulse wave errors in comparison with other methods. ConclusionsClassical noninvasive methods for estimating CABP were found to differ considerably in both the accuracy and error source, which provided theoretical evidence for understanding the specific advantages and disadvantages of each method. Knowledge about the method-specific error source and sensitivities of errors to different physiological/hemodynamic factors may contribute as theoretical references for interpreting clinical observations and exploring factors underlying large estimation errors, or provide guidance for optimizing existing methods or developing new methods.

The AllWISE Catalog as an Infrared Celestial Reference Frame in the Gaia Era

The launch of Gaia in 2013 December ushered in a new era of space astrometry, allowing for fundamental reference research at an unprecedented level of precision. The international celestial reference frames in the radio band and the Gaia celestial reference frame in the optical band have been established and are consistent within several microarcseconds for axes orientation. To bridge the gap between the visual and radio bands, an infrared reference frame should be investigated. We present a study aimed at constructing an infrared reference frame using the observations from the AllWISE catalog. It is compared with Gaia DR3 for approximately 0.57 million extragalactic sources and for a full set of 273 million sources. Systematic differences in positions and proper motions, such as magnitude or color equations and vector spherical harmonics, are derived. These systematic differences are comparable to the random errors of AllWISE measurements and can be used to improve the AllWISE source positions and proper motions, making the AllWISE catalog a valuable all-sky reference frame in the infrared band. Our investigation of the AllWISE catalog reveals that extragalactic sources and stars exhibit different astrometric properties. The global difference between the extragalactic source reference frame and the stellar reference frames is found to be 8.6 mas and 13.7 mas for global rotation and glide amplitudes, respectively. Such internal inconsistency should be considered when using AllWISE as an infrared reference frame. Finally, we determine the orientation of the mean Galactic plane using the calibrated source distribution of the AllWISE catalog.

Random measurement and prediction errors limit the practical relevance of two velocity sensors to estimate the 1RM back squat.

While maximum strength diagnostics are applied in several sports and rehabilitative settings, dynamic strength capacity has been determined via the one-repetition maximum (1RM) testing for decades. Because the literature concerned several limitations, such as injury risk and limited practical applicability in large populations (e.g., athletic training groups), the strength prediction via the velocity profile has received increasing attention recently. Referring to relative reliability coefficients and inappropriate interpretation of agreement statistics, several previous recommendations neglected systematic and random measurement bias. This article explored the random measurement error arising from repeated testing (repeatability) and the agreement between two common sensors (vMaxPro and TENDO) within one repetition, using minimal velocity thresholds as well as the velocity = 0m/s method. Furthermore, agreement analyses were applied to the estimated and measured 1RM in 25 young elite male soccer athletes. The results reported repeatability values with an intraclass correlation coefficient (ICC) = 0.66-0.80, which was accompanied by mean absolute (percentage) errors (MAE and MAPE) of up to 0.04-0.22m/s and ≤7.5%. Agreement between the two sensors within one repetition showed a systematic lower velocity for the vMaxPro device than the Tendo, with ICCs ranging from 0.28 to 0.88, which were accompanied by an MAE/MAPE of ≤0.13m/s (11%). Almost all estimations systematically over/ underestimated the measured 1RM, with a random scattering between 4.12% and 71.6%, depending on the velocity threshold used. In agreement with most actual reviews, the presented results call for caution when using velocity profiles to estimate strength. Further approaches must be explored to minimize especially the random scattering.

Stressing the Relevance of Differentiating between Systematic and Random Measurement Errors in Ultrasound Muscle Thickness Diagnostics

BackgroundThe majority of studies that explore changes in musculature following resistance training interventions or examine atrophy due to immobilization or sarcopenia use ultrasound imaging. While most studies assume acceptable to excellent reliability, there seems to be unawareness of the existing absolute measurement errors. As early as 1998, methodological research addressed a collective unawareness of the random measurement error and its practical indications. Referring to available methodological approaches, within this work, we point out the limited value of focusing on relative, correlation-based reliability indices for the interpretability in scientific research but also for clinical application by assessing 1,512 muscle thickness values from more than 400 ultrasound images. To account for intra- and inter-day repeatability, data were collected on two consecutive days within four testing sessions. Commonly-stated reliability values (ICC, CV, SEM and MDC) were calculated, while evidence-based agreement analyses were applied to provide the accompanied systematic and random measurement error.ResultsWhile ICCs in the range of 0.832 to 0.998 are in accordance with the available literature, the mean absolute percentage error ranges from 1.34 to 20.38% and the mean systematic bias from 0.78 to 4.01 mm (all p ≤ 0.013), depending on the measurement time points chosen for data processing.ConclusionsIn accordance with prior literature, a more cautious interpretation of relative reliability values should be based on included systematic and random absolute measurement scattering. Lastly, this paper discusses the rationale for including different measurement error statistics when determining the validity of pre-post changes, thus, accounting for the certainty of evidence.

The measurement of "high-impact chronic pain": Limitations and alternative methods.

Chronic pain is known to be an important construct in clinical practice and a particular form of chronic pain, high-impact chronic pain (HICP), has gained recent interest and attention by pain clinicians, epidemiologists, and clinical researchers. The purpose of our Topical Review is to describe the historical development of measures of HICP and to explore the psychometric properties of HICP as well as to present alternative measurement methods. We identified strengths and weaknesses of the psychometric characteristics of HICP measures. Limitations of existing HICP measures were discussed and summarized and alternatives to current methods were proposed. HICP operational definitions show variability across studies. All definitions cannot be correct, but which ones are incorrect cannot be determined as there is no gold standard. Random measurement error and recall bias are among the other limitations of current HICP measures. Model-based definitions of HICP, the discrete (for epidemiologic applications) and continuous (for clinical applications) latent variable models are discussed as likely superior alternatives to current methods. Limitations of existing HICP methods are discussed and alternative development approaches to HICP measures are presented. The use of either discrete or continuous latent variable models would improve upon the psychometric characteristics of current HICP evidence. Examples are used to illustrate the benefits of latent variable models over traditional observed variable conceptualizations as the measurement of HICP continues to develop. This work takes the position that current methods of measuring high impact chronic pain (HICP) likely contain substantial error. We have endorsed an alternative approach for several psychometrically grounded reasons. We recommend that future work consider the discrete latent variable framework for dichotomous measures of HICP and the continuous latent variable framework for continuous measures of HICP. The paper provides illustrative examples of these methods for a different patient reported measure that is lacking a gold standard, much like HICP measures.

Attention-Deficit/Hyperactivity Disorder Symptoms as a Function of Arousal and Executive Vigilance: Testing Halperin and Schulz’s Neurodevelopmental Model in a Sample of Community Adults

Halperin and Schulz’s neurodevelopmental model postulates that the onset of attention-deficit/hyperactivity disorder (ADHD) in childhood is due to subcortical alterations, whereas the disorder’s trajectory into adulthood depends on the development of executive functions. Based on a dimensional framework of ADHD, Coll-Martín et al. (2021) found support for the model in an adult community sample assessed on arousal and executive vigilance. The present study is a preregistered (https://osf.io/tkdq7) close replication of Coll-Martín et al. expressly aimed to test the two predictions of the model. A sample of college students (N = 292 valid; 49% women; 18–30 years, M = 21.7) from a Spanish university completed self-reports of ADHD symptoms in childhood (retrospectively) and adulthood and performed the online version of an attentional task (the ANTI-Vea). Our preregistered hypotheses achieved an acceptable statistical power for the effects of interest, even after accounting for random measurement error. Despite this, none of them replicated the findings of the original study: Only the unexpected negative correlation between executive vigilance and symptoms in childhood was significant, thereby not supporting the theoretical predictions. The lack of support for the dissociation pattern hypothesized by the neurodevelopmental model was robust to multiverse and exploratory analyses. At least in terms of vigilance, ADHD symptoms seem to share altered neurocognitive pathways across the lifespan, regardless of their time of onset. This challenges the notion of late-onset ADHD as a condition neuropsychologically distinct from child-onset ADHD. Future studies need to include complementary assessment methods and clinical groups.

A Comparison of Test-Retest Reliability and Practice Effects of Short Portable Mental State Questionnaire and Montreal Cognitive Assessment in Patients with Stroke.

To compare the test-retest reliabilities and minimal detectable change (MDC) of the Short Portable Mental State Questionnaire (SPMSQ) and the Montreal Cognitive Assessment (MoCA) in patients with stroke. 63 patients were recruited from 1 medical center. The SPMSQ and MoCA were administered twice, 2 weeks apart. Both measures showed high intraclass correlation coefficients (SPMSQ: 0.87; MoCA: 0.89) and acceptable MDC%s (SPMSQ: 14.8%; MoCA: 19.6%). A small correlation (r = 0.30) was found between the absolute difference and average in each pair of assessments in the SPMSQ, which was close to the criterion of heteroscedasticity. A small practice effect was observed in the MoCA (Cohen's d = 0.30). The SPMSQ demonstrated smaller random measurement error and an absence of practice effect. When comparing the psychometric properties of the SPMSQ and MoCA as outcome measures for assessing cognitive function in patients with stroke, the SPMSQ appears to be a more suitable choice than the MoCA.

Prospective longitudinal cohort study of quantitative muscle magnetic resonance imaging in a healthy control population.

Quantitative muscle magnetic resonance imaging (qMRI) is a valuable methodology for assessing muscular injuries and neuromuscular disorders. Notably, muscle diffusion tensor imaging (DTI) gives insights into muscle microstructural and macrostructural characteristics. However, the long-term reproducibility and robustness of these measurements remain relatively unexplored. The purpose of this prospective longitudinal cohort study was to assess the long-term robustness and range of variation of qMRI parameters, especially DTI metrics, in the lower extremity muscles of healthy controls under real-life conditions. Twelve volunteers (seven females, age 44.1 ± 12.1 years, body mass index 23.3 ± 2.0 kg/m2) underwent five leg muscle MRI sessions every 20 ± 4 weeks over a total period of 1.5 years. A multiecho gradient-echo Dixon-based sequence, a multiecho spin-echo T2-mapping sequence, and a spin-echo echo planar imaging diffusion-weighted sequence were acquired bilaterally with a Philips 3-T Achieva MR System using a 16-channel torso coil. Fifteen leg muscles were segmented in both lower extremities. qMRI parameters, including fat fraction (FF), water T2 relaxation time, and the diffusion metrics fractional anisotropy (FA) and mean diffusivity (MD), were evaluated. Coefficients of variance (wsCV) and intraclass correlation coefficients (ICCs) were calculated to assess the reproducibility of qMRI parameters. The standard error of measurement (SEM) and the minimal detectable change (MDC) were calculated to determine the range of variation. All tests were applied to all muscles and, subsequently, to each muscle separately. wsCV showed good reproducibility (≤ 10%) for all qMRI parameters in all muscles. The ICCs revealed excellent agreement between time points (FF = 0.980, water T2 = 0.941, FA = 0.952, MD = 0.948). Random measurement errors assessed by SEM and the MDC were low (< 12%). In conclusion, in this study, we showed that qMRI parameters in healthy volunteers living normal lives are stable over 18 months, thereby defining a benchmark for the expected range of variation over time.

Test of conformance or non-conformance with geometrical specifications

Abstract Systematic and random measurement errors are the cause of uncertain conformance and non-conformance statements. A wrong conformance statement occurs if the workpiece is accepted and is a reject part (type II error or false-negative) and a wrong non-conformance statement if the workpiece is rejected and is an in-spec part (type I error or false-positive). In order to avoid type I and type II errors, measurement uncertainty must be taken into account in the conformance and non-conformance testing. In practice, some procedures are used to consider measurement errors or the uncertainty that deviate from the state-of-the-art in research and technology. As these methods have become established over many years, they are still widely used despite better theoretical knowledge. The standard ISO 14253-1:2017 specifies a procedure based on probability and measurement uncertainty that is in accordance to the internationally accepted “Guide to the expression of uncertainty in measurement” and its supplements but is often not used due to the complexity of the evaluation of measurement uncertainty. In this contribution we give an overview and comparison of the different existing methods and provide an suggestion for supplementing the standard ISO 14253-1:2017, as Monte Carlo simulations enable a direct probability-based conformance and non-conformance testing even for complex measurement processes.

Characterizing Femtosecond Laser Tagging Performance With A 1030 Nm Yb:KGW Laser And Application To Continuous 50 KHz Velocimetry

This paper presents a comprehensive evaluation of unseeded Femtosecond Laser Electronic Excitation Tagging (FLEET) using a 1030 nm Yb:KGW femtosecond laser. In the first part of this work, fundamental studies were conducted in a vacuum chamber to assess the fluorescence process in dry air and nitrogen. Specifically, the effects of pressure, pulse energy, and focal length on the temporal decay of the nitrogen first and second positive band emission are quantified and reported. The results show similar trends to previous reports of FLEET emission behavior at other visible and near infra-red excitation wavelengths, indicating that the same fundamental fluorescence mechanism occurs after 1030 nm excitation. In the second part of this work, we explore the application of 1030 nm FLEET to continuous high-speed velocimetry in an under-expanded laboratory jet. Unseeded velocity measurements at 50 kHz in air were acquired at several points in the flow using burst gating of a high-speed intensifier. Frequency domain analysis of over 50,000 time-resolved measurements show strong periodic fluctuations near 8 kHz due to an acoustic resonance near the jet exit. Analysis of the streamwise velocity autocorrelation is shown to provide a means to separate uncorrelated random measurement error from true flow fluctuations. These results establish a first demonstration of continuous, high-speed FLEET measurements at 1030 nm and provides foundational data to inform future applications of this unseeded velocimetry technique in turbulent high-speed flows.

The use of the Isolation Forest model for anomaly detection in measurement data

The subject of the research is the Isolation Forest model, which is a powerful and efficient tool for detecting anomalies in measurement data and outliers, applicable in various fields where ensuring high accuracy and reliability of measurements is important. The goal of the study is to apply the Isolation Forest model to identify unusual or anomalous patterns that differ from typical patterns in the output data. This is achieved by isolating anomalous patterns from normal ones through the construction of multiple different decision trees. The task of the research is to detect outliers in data obtained during the preparation for international comparisons on the state primary standard for mass and volume flow rate of fluid, mass and volume of fluid flowing through a pipeline, by measuring with a сoriolis flowmeter. Data collected during metrological studies undergo processing by the model to detect anomalies. This model analyzes the data and identifies anomalous or outlier values that may indicate systematic or random measurement errors. It enables quick and efficient detection of even the smallest deviations in the data, helping to maintain high accuracy and reliability of measurement results. The main methods for detecting outliers in statistical analysis, which are distribution-independent, are the Grubbs' criterion, interquartile range distribution, and standard deviation. They are sensitive to sample size but are simple and understandable tools. However, the Isolation Forest model also has its limitations, particularly it can be resource-demanding for large datasets. Additionally, it is necessary to consider that using the model requires proper parameter tuning to achieve optimal results. The results of the research include assessment of the Isolation Forest model's effectiveness by comparing it with traditional outlier detection methods. Comparative analysis of the results of different approaches to the same task is an effective method for evaluating the model's performance. Conclusion. The article concludes with the perspective of further research development in this direction. The work will focus on further developing methods for detecting anomalies in measurement data and improving the accuracy and reliability of measurement results in various application fields, which can find broad applications in science and industry.

Optimal designs of accelerated degradation tests with random shock failures and measurement errors

AbstractAccelerated degradation tests (ADTs) are widely used for assessing the reliability of long‐life products. During an ADT, accelerated stresses not only expedite the degradation of test products but also increase the likelihood of encountering traumatic shocks. Moreover, it is important to acknowledge that measurement errors can be inevitable during the observation process of an ADT. Unfortunately, these errors are often overlooked in the optimal design of the ADT, especially when multiple competing failure modes are present. In this article, we propose a new approach to design ADTs when measurement errors exist and test products suffer from degradation failures and random shock failures. We utilize the Wiener process to model the degradation path, incorporating normally distributed measurement errors, and an exponential distribution to fit the time between random shock failures. Given the number of test products and the termination time, we optimize the ADT plans under three common design criteria. The equivalence theorem is used to verify the optimality of the optimal ADT plans. A real‐life example and sensitivity analysis are provided to illustrate our proposed method. The results demonstrate that when competing failure modes are present, the optimal ADT plans, which account for measurement errors, differ significantly from those that do not consider such errors.

Computing levels of nutrient inadequacy from household consumption and expenditure surveys: A case study

This paper presents an approach to estimate the between-subject variability in nutrient intake (through the coefficient of variation [CV]) and a method to estimate the prevalence of nutrient inadequacy (PoNI) (for eight micronutrients) using household consumption and expenditure survey (HCES) data. Prevalence values are compared to individual-level estimates derived using the National-Cancer-Institute method. Data come from the 2015 Bangladesh Integrated-Household-Survey, which conducted a household-level 7-day recall (7DR) and two rounds of individual-level 24-hour recall (24HR), filled by one respondent on behalf of all members, for the same rural households. The PoNI values based on 7DR are lower than those calculated from 24HR data, due to the larger average intake estimates from 7DR data. After controlling for differences in average intake estimates and adjusting household-level data for random measurement errors, the PoNI values from 7DR and 24HR data are remarkably close. This highlights the potential use of HCES data (conducted according to international agreed standards) for estimating the level of between-subject variability in usual nutrient intake in a population. The CVs from HCES could be used to compute the PoNI using average intake estimates from individual-level data; and the inadequacy of global nutrient supply using Supply and Utilization Accounts data.

Identification of Engine Thrust and Aerodynamic Drag Force According to Flight Test Data with Smoothing of Random Measurement Errors

Identification of Engine Thrust and Aerodynamic Drag Force According to Flight Test Data with Smoothing of Random Measurement Errors

Method for measuring non-stationary motion attitude based on MEMS-IMU array data fusion and adaptive filtering

The error characteristics of a typical commercial MEMS-IMU were analyzed. Using 15 ICM-42688 sensors, a 3*5 MEMS-IMU array was designed, and a weighted data fusion method based on Allan variance for the MEMS-IMU array was proposed. This effectively reduces the random measurement errors of the MEMS-IMU, providing a data foundation for the precise measurement of motion and attitude of robots, vehicles, aircraft, and other systems under low-cost conditions. The text describes a measurement method for non-stationary motion attitudes of sports vehicles based on adaptive Kalman-Mahony. Specifically, it first uses adaptive Kalman filter on array sensor data to calculate the measurement noise in real-time and adaptively adjust the filtering gain. Then, it determines the compensation coefficient of the accelerometer to the gyroscope angular velocity based on the motion state of the vehicle, and solves the attitude through complementary filtering to obtain the motion attitude quaternion. Finally, it converts it into the roll angle, pitch angle, and yaw angle of the sports vehicle. Experimental results show that the proposed MEMS-IMU array weighted data fusion method based on Allan variance has significant advantages over traditional single MEMS-IUM methods and traditional average weighting methods in reducing sensor angle random walk and zero drift instability. The proposed adaptive Kalman-Mahony attitude measurement method also shows a significant improvement in the accuracy of non-stationary motion attitude measurement compared to traditional methods.

Constructing predictive models for seismic oscillation parameters using covariance functions and Doppler effect phenomena: A case study of InSight mission V2 data

An ability to construct predictive models for identifying seismic oscillation parameters by using the mathematics of covariance functions and Doppler effect phenomena is examined in this work. In the calculations, the Mars seismic oscillations measurement data from InSight Mission V2, observed in the months May, June and July of 2019, was used. To analyze the observation data arrays the Doppler phenomena and the expressions of covariance functions were employed. The seismic oscillations trend's intensity vectors were assessed by least squares method, and the random errors of measurements at the stations were eliminated partially as well. The estimates of the vector's auto-covariance and cross-covariance functions were derived by altering the quantization interval on the general time scale while varying the magnitude of the seismic oscillation vector on the same time scale. To detect the mean values of z —the main parameter of Doppler expression— we developed a formula by involving the derivatives of cross-covariance functions of a single vector and algebraic sum of the relevant vectors.

U-SHAPED RELATIONSHIP BETWEEN BLOOD PRESSURE AND ALL-CAUSE MORTALITY IN OLDER ADULTS: THE SHIZUOKA STUDY

Objective: Not only high blood pressure (BP) but also low BP was suggested to be associated with all-cause mortality in older adults. However, the actual mechanism responsible for the association between low BP and mortality has not been identified. This study aimed to investigate a possible confounding of antihypertensive drug use, potential comorbidities, and functional disability, as well as random error in BP measurement, in the association between low BP and all-cause mortality. A possibility of reverse causation (i.e., lower BP was caused by cardiovascular frailty at the end of life) was also investigated. Design and method: We use the Kokuho Database in a local government, which is a prefecture-wide large dataset including data on annual health checkup and health and care insurance claims. This longitudinal observational study analyzed 337,988 individuals aged &amp;#8805; 65 (mean age: 73.5) years. The earliest day of participation in the annual health checkup during the study period from 2012 to 2020 was defined as baseline. Data on comorbidities, prescribed antihypertensive medications, functional disability levels, and incidence of stroke and all-cause mortality were obtained from the insurance claims database. Results: During a mean follow-up period of 5.3 years, there were 27,497 cases of all-cause mortality and 11,090 cases of stroke. A U-shaped association was observed between systolic BP and all-cause mortality in participants aged &amp;#8805; 75 years. The hazard ratio of systolic BP &lt;110 mmHg for all-cause mortality was 1.14 (95% confidence interval: 1.07–1.21, P &lt; 0.001). Meanwhile, the hazard ratio of BP &amp;#8805; 150 mmHg was 1.16. The U-shaped association remained significant in the subanalyses of individuals without severe comorbidities, antihypertensive drug use, or functional disability. Similar results were also observed in the analysis excluding early-mortality cases within 3 years after the baseline, as well as in the analysis using the average systolic BP measured during at least two different occasions within a 3-year period as a baseline value. In contrast, stroke risk increased linearly with increasing BP. Conclusions: Low BP was associated with all-cause mortality in older adults. The U-shaped association may not be attributed to the previously suggested factors.

Reduction of ADC bias in diffusion MRI with deep learning-based acceleration: A phantom validation study at 3.0 T

PurposeFurther acceleration of DWI in diagnostic radiology is desired but challenging mainly due to low SNR in high b-value images and associated bias in quantitative ADC values. Deep learning-based reconstruction and denoising may provide a solution to address this challenge. MethodsThe effects of SNR reduction on ADC bias and variability were investigated using a commercial diffusion phantom and numerical simulations. In the phantom, performance of different reconstruction methods, including conventional parallel (SENSE) imaging, compressed sensing (C-SENSE), and compressed SENSE acceleration with an artificial intelligence deep learning-based technique (C-SENSE AI), was compared at different acceleration factors and flip angles using ROI-based analysis. ADC bias was assessed by Lin's Concordance correlation coefficient (CCC) followed by bootstrapping to calculate confidence intervals (CI). ADC random measurement error (RME) was assessed by the mean coefficient of variation (CV¯) and non-parametric statistical tests. ResultsThe simulations predicted increasingly negative bias and loss of precision towards lower SNR. These effects were confirmed in phantom measurements of increasing acceleration, for which CCC decreased from 0.947 to 0.279 and CV¯ increased from 0.043 to 0.439, and of decreasing flip angle, for which CCC decreased from 0.990 to 0.063 and CV¯ increased from 0.037 to 0.508. At high acceleration and low flip angle, C-SENSE AI reconstruction yielded best denoised ADC maps. For the lowest investigated flip angle, CCC = {0.630, 0.771 and 0.987} and CV¯={0.508, 0.426 and 0.254} were obtained for {SENSE, C-SENSE, C-SENSE AI}, the improvement by C-SENSE AI being significant as compared to the other methods (CV: p = 0.033 for C-SENSE AI vs. C-SENSE and p < 0.001 for C-SENSE AI vs. SENSE; CCC: non-overlapping CI between reconstruction methods). For the highest investigated acceleration factor, CCC = {0.479,0.926,0.960} and CV¯={0.519,0.119,0.118} were found, confirming the reduction of bias and RME by C-SENSE AI as compared to C-SENSE (by trend) and to SENSE (CV: p < 0.001; CCC: non-overlapping CI). ConclusionADC bias and random measurement error in DWI at low SNR, typically associated with scan acceleration, can be effectively reduced by deep-learning based C-SENSE AI reconstruction.