Search

Deep neural networks trained on labeled medical data face major challenges owing to the economic costs of data acquisition through expensive medical imaging devices, expert labor for data annotation, and large datasets to achieve optimal model performance. The heterogeneity of diseases, such as Alzheimer's disease, further complicates deep learning because the test cases may substantially differ from the training data, possibly increasing the rate of false positives. We propose a reconstruction‐based self‐supervised anomaly detection model to overcome these challenges. It has a dual‐subnetwork encoder that enhances feature encoding augmented by skip connections to the decoder for improving the gradient flow. The novel encoder captures local and global features to improve image reconstruction. In addition, we introduce an entropy‐based image conversion method. Extensive evaluations show that the proposed model outperforms benchmark models in anomaly detection and classification using an encoder. The supervised and unsupervised models show improved performances when trained with data preprocessed using the proposed image conversion method. [ABSTRACT FROM AUTHOR]

A mathematical model is presented to study the cilia beating modulated radiated ternary nanofluid flow between two cilia‐carpeted walls propagating with an applicable phase difference to show the asymmetric nature of the pumping. To regulate further the cilia‐based pumping, electroosmosis, and magnetohydrodynamics mechanisms are considered in this model. Ternary nanofluid is considered with TiO2, SiO2, and Al2O3 nanoparticles dispersed in pure blood. The momentum slip condition is employed to derive the model solution. Thermal radiation and buoyancy effects are also taken into consideration. Mathematica NDSolve is utilized to simulate the numerical results for the velocity, temperature, and concentration profiles under the influence of the emerging parameters. Moreover, entropy generation and variation in Nusselt and Bejan numbers are computed for better thermal analysis. A comparative thermal analysis for unary, binary, and ternary nanofluids is also made. The results indicate that the trihybrid nanofluid offers an 8.5% increase in the heat transfer rate as compared to the conventional fluid. Furthermore, the Brownian motion is responsible for the enhancement in fluid temperature, while electric double layer thickness helps in increasing heat transfer irreversibility around the channel lower wall. It is found that the total entropy of the system at the channel upper and lower space is minimized in the case of ternary fluid rather than conventional fluid. The promising results of the present model reflect huge applications in biomedical engineering, particularly in designing thermal cilia‐based microfluidic devices. [ABSTRACT FROM AUTHOR]

An investigation is carried out on the analysis of entropy on the flow of non‐Newtonian fluid, in particular, micropolar fluid past an inclined channel. To enhance the fluid properties, velocity and thermal slip conditions are taken into consideration. At the outset, the novelty of the present investigation lies on the analysis of entropy generation that occurs due to the temperature differences between the media. The governing nonlinear equations are transformed to nonlinear ODE by the use of suitable transformed nondimensional variables. Furthermore, the motivation for the study is the solution of these governing equations using the semi‐analytical technique, namely, the variation parameter method. The behavior of the flow phenomena is characterized by the contributing parameters, in particular, the Bejan number, on the entropy are displayed via graphs and tables and elaborated in Section 5. The results reveal that the microrotation profile exhibited its dual character with an augmentation of the inclined angle, and both the coupling parameter and the Reynolds number are favorable in resisting entropy. [ABSTRACT FROM AUTHOR]

In this study, the electrically conducting nanoliquid film on a slippery inclined heated surface with convective cooling under the effect of the reactive and radiating magnetic field is numerically investigated. The nonlinear constructing governing differential equations are then converted into a set of dimensionless ordinary differential equations by similarity variables, and the systems are numerically solved by shooting method together with the Runge‐Kutta method. Some of the parameters involved in the flow system are then examined graphically and discussed with respect to the velocity, temperature, entropy generation rate, skin friction, Nusselt number, and the Bejan number profile. [ABSTRACT FROM AUTHOR]

Food quality ratings can be used for product differentiation and strategic marketing. This article develops a differentiation measure based on the cross-entropy of two distributions for speciality coffees. It examines whether quality ratings achieve their objectives for speciality coffee supply chains. The estimation, using e-auctions data, indicates that the quality rating is more informative in the Cup of Excellence (CofE) relative to the Q auction. To determine significance, a statistical test for comparing the information measure across different e-auctions and segments of the quality rating was developed. The significant difference between the two auctions is not a surprise as the auctions trade coffee lots of different volumes and the quality evaluation is more important in the case of the CofE, with its jury of renowned coffee experts. The total cross-entropy measure of product differentiation shows that the 100-point quality rating does not carry the same information content and meaning in the two different e-auctions and their respective supply chains. The cross-entropy measure of product differentiation is shown to be a good indicator of the information and the value created throughout the speciality food supply chain. [ABSTRACT FROM AUTHOR]

Entropy (ENT) is a newly developed measure of the complexity of heart rate variability (HRV). The aim of this study was to characterize the complexity of HRV in patients with end-stage renal disease (ESRD) and to find a possible clinical utility. Healthy subjects and patients with ESRD undergoing hemodialysis (HD) were recruited. The HD population consisted of patients with and without diabetes mellitus (DM). An electrocardiogram was recorded before HD, and blood pressure was measured during HD. The coefficients of variation of R-R intervals, high- and low-frequency components, and ratio of the low- to high-frequency components were measured as variables of HRV. The ENT was used to describe the complexity of HRV. Forty-six healthy subjects and 27 HD patients participated in this study. The ENT negatively correlated with the duration of DM (p = 0.001), systolic blood pressure (p = 0.003), and mean blood pressure (p = 0.004) before a HD session. ENT in HD patients was lower than that in healthy subjects (p < 0.01). ENT in HD patients with DM was lower than that in HD patients without DM (p < 0.01). The change in systolic blood pressure (ΔSBP) during a HD session showed high correlations to ENT and ultrafiltration rate (UFR) of the dialyzer. The following equation was obtained: ΔSBP = 2.25 × ENT − 2.28 × UFR − 21.27 (R2 = 0.805; p < 0.0001). ENT decreased with uremic and diabetic status. ENT also represents a possible prediction of hypotension during a HD session. [ABSTRACT FROM AUTHOR]