Abstract

Abstract

Latest developments in high-strength Magnetic Resonance Imaging (MRI) scanners with in-built high resolution, have dramatically enhanced the ability of clinicians to diagnose tumours and rare illnesses. However, their high-strength transient magnetic fields induce unwanted eddy currents in shielding components, which result in fast vibrations, noise, imaging artefacts and, ultimately, heat dissipation, boiling off the helium used to super-cool the magnets. Optimum MRI scanner design requires the capturing of complex electro-magneto-mechanical interactions with high fidelity computational tools. During production cycles, this is known to be extremely expensive due to the large number of configurations that need to be tested. There is an urgent need for the development of new cost-effective methods whereby previously performed computations can be assimilated as training solutions of a surrogate digital twin model to allow for real-time simulations. In this paper, a Reduced Order Modelling technique based on the Proper Generalised Decomposition method is presented for the first time in the context of MRI scanning design, with two distinct novelties. First, the paper derives from scratch the offline higher dimensional parametrised solution process of the coupled electro-magneto-mechanical problem at hand and, second, a regularised adaptive methodology is proposed for the circumvention of numerical singularities associated with the ill-conditioning of the discrete system in the vicinity of resonant modes. A series of numerical examples are presented in order to illustrate, motivate and demonstrate the validity and flexibility of the considered approach.

Abstract

Abstract

'''Diabetic retinopathy, a type of eye disease, is a danger for diabetics (DR). Manual labour, which is prone to inaccuracy and time consuming, makes dealing with this illness considerably more difficult. Normally computer-assisted diagnosis has appeared as a promising tool for the early identification and severity grading of DR. As technologies are revolutionizing day by day, in which the most advance technology deep learning's algorithm gives a tremendous support for healthcare fields. In this article propose an efficient classification of DR models such as Generative Adversarial Network with Deep Belief Network (GAN-DBN), this category the DR into different grades and to identify the severity. Obtaining medical data is also a tedious and challenging one because it is quite expensive and confidential, to overcome this the proposed model is act as a data augmentation model, moreover the features in the input data’s are reduced by Dimensionality reduction Module (DRM) based on Principal Component Analysis. Enhancing retinal fundus images using a median filter, morphological operation, and histogram equalization approaches. When compared to previous classification models, experimental results on fundus images using the MESSIDOR database reveal that the GAN-based DBN model obtained good processing performances with 99.2% accuracy in the suggested classification task.