Search

Background The prognostic role of the cardiothoracic ratio (CTR) in chronic kidney disease (CKD) remains undetermined. Methods We conducted a retrospective cohort study of 3117 patients with CKD aged 18–89 years who participated in an Advanced CKD Care Program in Taiwan between 2003 and 2017 with a median follow up of 1.3(0.7–2.5) and 3.3(1.8–5.3) (IQR) years for outcome of end-stage renal disease (ESRD) and overall death, respectively. We developed a machine learning (ML)–based algorithm to calculate the baseline and serial CTRs, which were then used to classify patients into trajectory groups based on latent class mixed modelling. Association and discrimination were evaluated using multivariable Cox proportional hazards regression analyses and C-statistics, respectively. Results The median (interquartile range) age of 3117 patients is 69.5 (59.2–77.4) years. We create 3 CTR trajectory groups (low [30.1%], medium [48.1%], and high [21.8%]) for the 2474 patients with at least 2 CTR measurements. The adjusted hazard ratios for ESRD, cardiovascular mortality, and all-cause mortality in patients with baseline CTRs ≥0.57 (vs CTRs P = 0.04) for cardiovascular mortality and 0.697 (vs 0.693, P Conclusions Our findings support the real-world prognostic value of the CTR, as calculated by a ML annotation tool, in CKD. Our research presents a methodological foundation for using machine learning to improve cardioprotection among patients with CKD.

Background: The prognostic role of the cardiothoracic ratio (CTR) in chronic kidney disease (CKD) remains undetermined. Methods: We conducted a retrospective cohort study of 3117 patients with CKD aged 18 to 89 years who participated in an Advanced CKD Care Program in Taiwan between 2003 and 2017 with a median follow up of 1.3(0.7-2.5) and 3.3(1.8-5.3) (IQR) years for outcome of end-stage renal disease (ESRD) and overall death, respectively. We developed a machine learning (ML)–based algorithm to calculate the baseline and serial CTRs, which were then used to classify patients into trajectory groups based on latent class mixed modelling. Association and discrimination were evaluated using multivariable Cox proportional hazards regression analyses and C-statistics, respectively. Results: Here we show the median (interquartile range) age of 3117 patients is 69.5 (59.2-77.4) years. We create 3 CTR trajectory groups (low [30.1%], medium [48.1%], and high [21.8%]) for the 2474 patients with at least 2 CTR measurements. The adjusted hazard ratios for ESRD, cardiovascular mortality, and all-cause mortality in patients with baseline CTRs ≥0.57 (vs CTRs P = .04) for cardiovascular mortality and 0.697 (vs 0.693, P  Conclusions: Our findings support the real-world prognostic value of the CTR, as calculated by a ML annotation tool, in CKD. Our research presents a methodological foundation for using machine learning to improve cardioprotection among patients with CKD.

Glycosylated hemoglobin (HbA1c) targets for patients with chronic kidney disease (CKD) and type 2 diabetes remain controversial. To evaluate whether baseline HbA1c and HbA1c trajectories are associated with the risk of end-stage kidney disease (ESKD) and all-cause mortality, we recruited adult patients with CKD and type 2 diabetes from a “Pre-ESKD Program” at a medical center in Taiwan from 2003 to 2017. Group-based trajectory modeling was performed to identify distinct patient groups that contained patients with similar longitudinal HbA1c patterns. Cox proportional hazard models were used to estimate hazard ratios (HRs) of ESKD and mortality associated with baseline HbA1c levels and HbA1c trajectories. In the analysis related to baseline HbA1c (n = 4543), the adjusted HRs [95% confidence interval (CI)] of all-cause mortality were 1.06 (0.95–1.18) and 1.25 (95% CI, 1.07–1.46) in patients with an HbA1c level of 7%–9% (53–75 mmol/mol) and >9% (>75 mmol/mol), respectively, as compared with those with an HbA1c level < 7% ((n = 2692), three distinct longitudinal HbA1c trajectories were identified: nearly optimal (55.9%), moderate to stable (34.2%), and poor control (9.9%). Compared with the “nearly optimal” HbA1c trajectory group, the “moderate-to-stable” group did not have significantly higher mortality, but the “poorly controlled” group had 35% higher risk of mortality (adjusted HR = 1.35, 95% CI = 1.06–1.71). Neither baseline levels of HbA1c nor trajectories were associated with ESKD risk. In conclusion, in patients with CKD and type 2 diabetes, poor glycemic control was associated with an elevated risk of mortality but not associated with a risk of progression to ESKD.

The fasting blood glucose (FBG) values extracted from electronic medical records (EMR) are assumed valid in existing research, which may cause diagnostic bias due to misclassification of fasting status. We proposed a machine learning (ML) algorithm to predict the fasting status of blood samples. This cross-sectional study was conducted using the EMR of a medical center from 2003 to 2018 and a total of 2,196,833 ontological FBGs from the outpatient service were enrolled. The theoretical true fasting status are identified by comparing the values of ontological FBG with average glucose levels derived from concomitant tested HbA1c based on multi-criteria. In addition to multiple logistic regression, we extracted 67 features to predict the fasting status by eXtreme Gradient Boosting (XGBoost). The discrimination and calibration of the prediction models were also assessed. Real-world performance was gauged by the prevalence of ineffective glucose measurement (IGM). Of the 784,340 ontologically labeled fasting samples, 77.1% were considered theoretical FBGs. The median (IQR) glucose and HbA1c level of ontological and theoretical fasting samples in patients without diabetes mellitus (DM) were 94.0 (87.0, 102.0) mg/dL and 5.6 (5.4, 5.9)%, and 92.0 (86.0, 99.0) mg/dL and 5.6 (5.4, 5.9)%, respectively. The XGBoost showed comparable calibration and AUROC of 0.887 than that of 0.868 in multiple logistic regression in the parsimonious approach and identified important predictors of glucose level, home-to-hospital distance, age, and concomitantly serum creatinine and lipid testing. The prevalence of IGM dropped from 27.8% based on ontological FBGs to 0.48% by using algorithm-verified FBGs. The proposed ML algorithm or multiple logistic regression model aids in verification of the fasting status.

International Classification of Diseases (ICD) code-based claims databases are often used to study infective endocarditis (IE). However, the quality of ICD coding can influence the reliability of IE research. The impact of complementing the ICD-only approach with data extracted from electronic medical records (EMRs) has yet to be explored.We selected the information of adult patients with discharge ICD codes for IE (ICD-9: 421, 112.81, 036.42, 098.84, 115.04, 115.14, 115.94, 424.9; ICD-10: I33, I38, I39) during 2005-2016 in China Medical University Hospital. Data extraction was conducted on the basis of the modified Duke criteria to establish a reference group comprising patients with definite or possible IE. Clinical characteristics and in-hospital mortality were compared between ICD-identified and Duke-confirmed cases. The positive predictive value (PPV) was used to quantify the IE identification performance of various phenotyping algorithms.A total of 593 patients with discharge ICD codes for IE were identified, only 56.7% met the modified Duke criteria. The crude in-hospital mortality for Duke-confirmed and Duke-rejected IE were 24.4% and 8.2%, respectively. The adjusted in-hospital mortality for ICD-identified IE was lower than that for Duke-confirmed IE by a difference of 5.1%. The best PPV was achieved (0.90, 95% CI 0.86-0.93) when major components of the Duke criteria (positive blood culture and vegetation) were integrated with ICD codes.Integrating EMR data can considerably improve the accuracy of ICD-only approaches in phenotyping IE, which can improve the validity of EMR-based studies and their applications, including real-time surveillance and clinical decision support.

Poor sleep quality and insomnia are pervasive among patients with advanced chronic kidney disease (CKD); however, these health issues have not been systematically evaluated.Systematic review and meta-analysis.Adult patients with CKD not receiving kidney replacement therapy (KRT), as well as adults receiving KRT, including hemodialysis, peritoneal dialysis, and kidney transplantation.A systematic literature search using PubMed, Embase, and PsycNET, was conducted for articles published between January 1, 1990, and September 28, 2018.Data on the prevalences of poor sleep quality and insomnia in patients with CKD, including those receiving and not receiving KRT, were extracted.Pooled prevalences were estimated using a random-effects meta-analysis and were stratified according to age, CKD stage, World Health Organization region, risk of bias, Pittsburgh Sleep Quality Index score, and the different criteria for insomnia that were used at diagnosis.Of 3,708 articles, 93 were selected, and significant methodological heterogeneity was present. The pooled prevalences of poor sleep quality for CKD without KRT, hemodialysis, peritoneal dialysis, and kidney transplantation were 59% (95% CI, 44%-73%), 68% (95% CI, 64%-73%), 67% (95% CI, 44%-86%), and 46% (95% CI, 34%-59%), respectively. The corresponding prevalences of insomnia were 48% (95% CI, 30%-67%), 46% (95% CI, 39%-54%), 61% (95% CI, 41%-79%), and 26% (95% CI, 9%-49%), respectively. Insomnia was significantly more prevalent among patients aged 51-60 years and those aged60 years than among those aged 50 years. The prevalence of insomnia in the European region was the lowest of all World Health Organization regions.High interstudy heterogeneity.Approximately half of the patients with advanced CKD had poor sleep quality or insomnia, and the prevalence was even higher among those who received KRT. Kidney transplantation may reduce the burden of poor sleep quality and insomnia.

The World Health Organization (WHO) predicts that depression disorders will be widespread in the next 20 years. These disorders may affect a person’s general health and habits such as altered sleeping and eating patterns in addition to their interpersonal relationships. Early depression detection and prevention therefore becomes an important issue. To address this critical issue, we recruited 1453 individuals who use Facebook frequently and collected their Facebook data. We then propose an automatic depression detection approach, named Deep Learning-based Depression Detection with Heterogeneous Data Sources (D3-HDS), to predict the depression label of an individual by analyzing his/her living environment, behavior, and the posting contents in the social media. The proposed method employs Recurrent Neural Networks to compute the posts representation of each individual. The representations are then combined with other content-based, behavior and living environment features to predict the depression label of the individual with Deep Neural Networks. To our best knowledge, this is the first attempt that simultaneously considers all the content-based, behavior, and living environment features for depression detection. The experiment results on a real dataset show that the performance of our approach significantly outperforms the other baselines.

This work aimed to investigate the remediation performance and reaction behavior of the injection of nanoscale Fe 3 O 4 slurry coupled with the electrokinetic (EK) process for remediation of NO 3 − in a saturated soil with the texture of sandy clay. First, 0.8 wt% of polyacrylic acid (PAA) was added to the laboratory-prepared nanoscale Fe 3 O 4 to yield the nanoscale Fe 3 O 4 slurry. Thus prepared slurry was then injected into the anode reservoir of the EK set-up for remediation of nitrate-contaminated soil ([NO 3 − ] = 69.39–71.65 mg/kg) in the soil compartment ( L : 25 cm; Φ : 10 cm). Application of an electric field of 1 V/cm and daily injection of nanoscale Fe 3 O 4 slurry at a dose of 5 g/L were practiced during the test period of 12 d. The remediation results indicated that residual NO 3 − –N concentration of 1.35 mg/L in the anode reservoir was found to be lower than Taiwan EPA's Pollution Control Standards for Type I Groundwater Quality. In addition, a very low residual NO 3 − concentration in soil was detected. It indicates that the hybrid technology employed in this work is effective for remediation of nitrate in the subsurface following the adsorptive reduction model. The relevant reaction behavior was discussed.

The dynamic equations of a robot manipulator, derived from the Euler or Lagrangian formulation, are governed by a set of highly non-linear and strongly coupled differential equations. The design of a controller to ensure fast, smooth and accurate control of a robot m anipulator motion becomes a great challenge. This pa er discusses several methods of designing a controller for a robot manipulator. Performances of time-invariant and time-varying controllers designed by the computed torque technique and the feedback linearization technique are compared in the control of a 3-joint manipulator. It is noted that timevaryin controllers are superior to time-invariant contro!Llers in convergence time, steady state accuracy and control e fforts.

It is shown that any linear time-varying system can be transformed into a time-invariant one provided that its state transition matrix φ(t. t0) is known. In this paper two fairly large classes of linear time-varying systems that can be explicitly transformed into time-invariant ones without using full information on φ(t. t0) are identified. They are the algebraically invariable systems and the γ-algebraically invariable systems. These classes include the well known Floquet systems and the Euler systems as special cases. It is also shown that any commutative system is γ-algebraieally invariable. Explicit methods of finding the desired algebraic transformation and the tγtransformation are also given. Several examples are presented and their stability is discussed.

In this paper a newly developed theory concerning the analysis of the trajectory sensitivity of linear uncertain systems is first reviewed. An index suggested by the theory is then introduced to represent the sensitivity of a system. Based on this index, we devise a method for synthesizing low-sensitivity constant-gain feedback regulators. Both academic illustrative and practical design examples are provided to explain these concepts and techniques.

In the recent research, we have shown the intimate relationship between the state trajectory sensitivity of a linear uncertain system and the normality of the system matrix. Utilizing this property, we introduce two different indices of the normality of matrices as the basis and propose two new methods for the synthesis of low sensitivity linear state feedback regulators. In addition to the description of the background theory, the two new methods are respectively applied to practical and academic design examples for illustration. Finally, computer simulations verify the effectiveness of the design methods.

A nonlinear plant with parameter uncertainty is imbedded in a feedback system subjected to a finite set of inputs {x. (t)}. For each i there is a specified set of response tolerances. The synthesis procedure guarantees the response tolerances are satisfied over the range of uncertainty, for a large class of plants. The nonlinear plant is converted into an equivalent linear, time-invariant plant with parameter uncertainty, for which exact design is possible. Schauder’s fixed point theorem proves the equivalence is valid. The technique is applicable to any structure for which the equivalent linear, time-invariant problem is solvable.

This paper presents some explicit results on solution, stability, and transformation of a fairly broad class of linear time-varying systems. It is shown that for this special class of linear time-varying systems, the solution can be represented as a product of two matrix exponential functions and the system stability can be determined directly from eigenvalues of two constant matrices. Furthermore the system can be reduced to a linear time-invariant system by successive applications of an algebraic transformation and a t?? transformation. The generalized results given here contain several previously reported results as special Cases.

A new approach using the decomposition of A(t) to study the commutative class of linear time-varying systems is presented. A simpler and computer-implemental criterion of checking the commutativity of A(t) with its integral is given. More transparent results on solution, stability, decoupling, the controlability and observability are obtained. With the new approach, a commutative linear time-varying system can be studied in the same manner as that of a linear time-invariant system.

A successive decomposition method for solving linear time-varying systems is presented. With good insight on solvable classes, this method may become one of the most general approaches for solving larger classes of linear time-varying systems. An example is given to show wide varieties of decomposition one can use to obtain the solution.

Explicit results on solution, stability and reduction of a class of discrete-time linear time-varying systems are presented. It is shown that the state transition matrix can be computed explicitly by φ(k,0)=A1 kA2 k ,where A 1and A 2are constant matrices. The stability is completely determined by eigenvalues of A 1 and A 2. It is also shown that the system can be reduced to a time-invariant one by an algebraic transformation. Examples are given to show that stability criteria of linear time-invariant systems cannot, in general, be applied to time-varying cases.

Several classes of linear time-varying systems are found to be solvable, and explicit methods for obtaining their solutions are available. In this paper it will be shown that there exists an algebraic transformation such that any linear time-varying system can be transformed into any desired solvable class. Consequently, it is concluded that solvability of linear time-varying systems is only a property of the system representation. It is definitely not an inherent property of the dynamic system.

This paper presents theoretical methods which may have practical applications in determining the complete controllability of linear dynamical systems when the system A matrix, the input B matrix, and the output C matrix may be unknown. The criteria are obtained solely from the use of the input-state and the input-output experimental data of the given system.

In this paper a generalized definition of passivity is introduced. Sufficient conditions of passivity for a certain class of time-varying non-linearities associating with various RL, RC and RLC multipliers are presented. It is pointed out that these passivity results can be used in conjunction with a Popov-type frequency domain condition to obtain many very general stability criteria of non-linear time-varying feedback systems in a systematic and unified manner by means of the passivity theorem.

The stability of a single-input, single-output, singleloop, linear, time-invariant system is related to the properties of its open-loop gain. The impulse response of the open-loop system may be of the form g(t) = r + g_{a}(t) + \sum_{i=0}^{\infty} g_{i} \delta (t - t_{i}) where r is a nonnegative constant, g_{a} is integrable on [0, \infty) , and \sum_{i=0}^{\infty} |g_{i}| \infty . If the Nyquist diagram of the open-loop gain does not go through nor encircle the critical point, then the closed-loop system is inputoutput stable, in the several meanings explained in the paper.

This correspondence presents some new results for linear time-varying systems \dot{x}(t)=A(t)x(t) . These new results include, 1) an explicit necessary and sufficient condition of stability that can be determined directly by joint eigenvalues of two constant matrices, and 2) a new class of reducible systems which needs not be periodic, as is required in Floquet theory.

A new concept of "eigenvalues" and "eigenvectors" of A(t) is introduced. This new concept reduces to the conventional one as a special case when A(t) has constant "eigenvectors." Its applications to algebraic transformation and diagonal realization are given.

In this paper, several solvable classes of linear time-varying systems are identified and explicit methods for finding their solutions are given. It is also shown that any linear time-varying system can be transformed into an algebraically equivalent commutative lineartime-varying system or a time-invariant system, provided that an appropriate nonsingular(time-varying)algebraic transformation can be found. Consequently it is concluded that the commutativity of a linear time-varying system is not an inherent property of a dynamic system, but rather is a representative property.

This paper presents two easy-to-check criteria that test respectively the necessary condition for the asymptotic stability and the sufficient condition for the instability of linear time-varying systems. A necessary and sufficient condition, defined in terms of the newly introduced concept of the mode-vectors is also given for stability of linear time-varying systems. An algorithm for finding the mode-vectors is presented. Several examples are given to illustrate the applications of the results in the paper.

The class of linear time-varying systems \dot{x}(t) = A (t)x(t) where the state transition matrix can be computed by \Phi(t,0) = exp (A_{1}t) exp (A_{2}t) is extended to the cases where eigenvalues of A(t) need not be independent of t .

Sufficient conditions for the stability of a class of multiplicative time-varying as well as time-invariant nonlinear systems are presented.

An example is given to show that even if the eigenvalues of the system A -matrix A(t) of a linear time-varying system \dot{x}(t) = A(t)x(t) are independent of t and some of them have positive real parts, the system is asymptotically stable.