An Approach Based on Mutually Informed Neural Networks to Optimize the Generalization Capabilities of Decision Support Systems Developed for Heart Failure Prediction

Available clinical methods for heart failure (HF) diagnosis are expensive and require a high-level of experts intervention. Recently, various machine learning models have been developed for the prediction of HF where most of them have an issue of over-fitting. Over-fitting occurs when machine learning based predictive models show better performance on the training data yet demonstrate a poor performance on the testing data and the other way around. Developing a machine learning model which is able to produce generalization capabilities (such that the model exhibits better performance on both the training and the testing data sets) could overall minimize the prediction errors. Hence, such prediction models could potentially be helpful to cardiologists for the effective diagnose of HF. This paper proposes a two-stage decision support system to overcome the over-fitting issue and to optimize the generalization factor. The first stage uses a mutual information based statistical model while the second stage uses a neural network. We applied our approach to the HF subset of publicly available Cleveland heart disease database. Our experimental results show that the proposed decision support system has optimized the generalization capabilities and has reduced the mean percent error (MPE) to 8.8% which is significantly less than the recently published studies. In addition, our model exhibits a 93.33% accuracy rate which is higher than twenty eight recently developed HF risk prediction models that achieved accuracy in the range of 57.85% to 92.31%. We can hope that our decision support system will be helpful to cardiologists if deployed in clinical setup.     

Micro-finite element simulation of trabecular-bone post-yield behaviour – effects of material model,element size and type

Micro-finite element (micro-FE) analysis became a standard tool for the evaluation of trabecular bone mechanical properties. The accuracy of micro-FE models for linear analyses is well established. However, the accuracy of recently developed nonlinear micro-FE models for simulations of trabecular bone failure is not known. In this study, a trabecular bone specimen was compressed beyond the apparent yield point. The experiment was simulated using different micro-FE meshes with different element sizes and types, and material models based on cortical bone. The results from the simulations were compared with experimental results to study the effects of the different element and material models. It was found that a decrease in element size from 80 to 40 μm had little effect on predicted post-yield behaviour. Element type and material model had significant effects. Nevertheless, none of the established material models for cortical bone were able to predict the typical descent in the load-displacement curve seen during compression of trabecular bone.     

Physiological control of an in-series connected pulsatile VAD: numerical simulation study

This paper investigates ventricular assist device (VAD)-assisted cardiovascular dynamics under proportion–integration–differentiation (PID) feedback control. Previously, we have studied the cardiovascular responses under the support of an in-series connected reciprocating-valve VAD through numerical simulation, and no feedback control was applied in the VAD. In this research, we explore the contribution of the VAD control on the circulatory dynamics assisted by the reciprocating-valve VAD, in response to the changing physiological conditions. The classical PID control algorithm is implemented to regulate the VAD stroke beat-to-beat, based on the error signal between the expected and the realistic mean aortic pressures. Simulation results show that under the PID VAD control, physiological variables such as left atrial, ventricular and systemic arterial pressures, cardiac output and ventricular volumes are satisfactorily maintained in the physiological ranges. With the online PID feedback control, operation of the reciprocating-valve VAD can be satisfactorily regulated to accommodate metabolic requirements under various physiological conditions including normal resting and exercise situations.     

Finite element analysis of Stryker Xia pedicle screw in artificial bone samples with and without supplemental cement augmentation

A validated, using in vitro biomechanical testing, finite element model was used to evaluate the affects of (1) cement augmentation and (2) an intact posterior cortex in osteoporotic bone. The presence of augmentation and/or a posterior cortical cortex increased the stabilization of the pedicle screw 2–5 fold. Placement of cement influenced failure load and toggle; with distal placement having the largest increase in failure load and decrease in cephalad–caudad toggle. The presence of posterior cortex caused a decrease in the amount of toggle, a proximal shift of the center of rotation and an increase in the maximum failure force.     

The mechanism of RNA oxidation involved in the development of heart failure

Heart failure (HF) has become a global public health problem due to its unclear pathogenesis. Our previous studies have found that RNA oxidation is associated with the occurrence and development of a variety of chronic diseases in the elderly, but whether RNA oxidation is related to the pathogenesis of HF remains unclear. Male Dahl salt-sensitive rats (DSSR) were divided into 8% NaCl groups and 0.3% NaCl groups. The blood pressure of DSSR, HE staining of cardiac tissue, cardiac function index of colour Doppler echocardiography and plasma N-terminal probrain Natriuretic Peptide (NT-ProBNP) were used to evaluate the model making. The levels of 8-hydroxyguanosine (8-oxoGsn) and 8-hydroxydeoxyguanosine (8-oxodGsn) in myocardium and urine of DSSR were determined by high-performance liquid chromatography–mass spectrometry (LC-MS/MS). The expression of ERK-MAPK pathway and MTH1 was detected by Western blot (WB). Rats in the 8% NaCl group developed heart failure symptoms such as increased blood pressure, myocardial hypertrophy, decreased diastolic function, and increased plasma NT-ProBNP. The content of 8-oxoGsn in urine and heart tissue also increased, which was positively correlated with the related indicators of heart failure. This process is also accompanied by the sequential activation of ERK-MAPK pathway molecules and the increase of MTH1. The mechanism of RNA oxidation and inhibition is related to the occurrence and development of HF, which may be involved through ERK-MAPK pathway.     

Cardiac and neuroprotection regulated by α1-adrenergic receptor subtypes

Sympathetic nervous system regulation by the α₁-adrenergic receptor (AR) subtypes (α_1A, α_1B, α_1D) is complex, whereby chronic activity can be either detrimental or protective for both heart and brain function. This review will summarize the evidence that this dual regulation can be mediated through the different α₁-AR subtypes in the context of cardiac hypertrophy, heart failure, apoptosis, ischemic preconditioning, neurogenesis, locomotion, neurodegeneration, cognition, neuroplasticity, depression, anxiety, epilepsy, and mental illness.     

Delayed Treatment Effects of Xanthine Oxidase Inhibition on Systolic Overload-Induced Left Ventricular Hypertrophy and Dysfunction

The nonpurine selective xanthine oxidase (XO) inhibitor febuxostat attenuates development of left ventricular (LV) hypertrophy and dysfunction in mice when treatment is initiated within 1 hour of transverse aortic constriction (TAC). This study investigated whether a 7-day delay of treatment with the XO inhibitors febuxostat or allopurinol would reverse TAC-induced changes after onset of heart failure (HF). Neither treatment significantly affected TAC-induced LV hypertrophy; only febuxostat caused a modest improvement in LV function (～10% increase in LV ejection fraction). However, the purine analog allopurinol tended to increase mortality compared with vehicle or febuxostat in HF mice.     

Oxidative stress and ischemia-modified albumin in chronic ischemic heart failure

Abstract

Objectives

Knowledge about the role of oxidative stress in human diseases, including cardiovascular system disorders, emphasizes the need for reliable markers of oxidative stress. Here, we evaluated the levels of the novel marker ischemia-modified albumin (IMA), albumin-adjusted IMA (adj-IMA), and the IMA/serum albumin ratio (IMAR) in patients with chronic ischemic heart failure (CIHF).

Methods

A total of 55 patients with CIHF and 40 age- and sex-matched healthy individuals were included in the study. Serum levels of IMA, total antioxidant status, and total oxidant status were analyzed, and the adj-IMA level, IMAR, and oxidative stress index were calculated.

Results

Serum IMA, IMAR, total oxidant status levels, and oxidative stress index were significantly higher in patients with CIHF than in the controls (all P < 0.0001), whereas albumin and total antioxidant status levels were significantly lower in the CIHF patients (P < 0.0001 and P = 0.0004, respectively). However, there was no significant difference in serum adj-IMA levels between the groups (P = 0.8).

Discussion

We observed impaired oxidant/antioxidant status in favor of oxidative stress in CIHF patients. Oxidative stress may be a key factor in the development of hypoalbuminemia in CIHF. Further studies are needed to establish the relationships among IMA, albumin, and redox balance in CIHF.