International Journal of Electrical and Electronics Research (IJEER) Open Access | Rapid and quality publishing Research Article | Volume 10, Issue 2 | Pages 312-319 | e-ISSN: 2347-470X 312 Website: www.ijeer.forexjournal.co.in A Novel Medical Image Segmentation Model ░ ABSTRACT- In deep learning-based computing vision for image processing, image segmentation is a prominent issue. There is promising generalisation performance in the medical image segmentation sector for approaches using domain generalisation (DG). Single domain generalisation (SDG) is a more difficult problem than conventional generalisation (DG), which requires numerous source domains to be accessible during network training, as opposed to conventional generalisation (DG). Color medical images may be incorrectly segmented because of the augmentation of the full image in order to increase model generalisation capacity. An arbitrary illumination SDG model for improving generalisation power for colour image segmentation approach for medical images through synthesizing random radiance charts is presented as a first solution to this challenge. Color medical images may be decomposed into reflectivity and illumination maps using retinex-based neural networks (ID-Nets). In order to provide medical colour images under various lighting situations, illumination randomization is used to enhance illumination maps. A new metric, TGCI, called the transfer gradient consistency index was devised to quantify the performance of the breakdown of retinal images by simulating physical lighting. Two of the existing retinal image segmentation tasks are tested extensively in order to assess our suggested system. According to the Dice coefficient, our framework surpasses previous SDGs and image improvement algorithms, outperforming the best SDGs by up to 1.7 per cent. Keywords: Image segmentation, medical data, machine learning, computer vision, domain generalisation. ░ 1. INTRODUCTION Computer-aided diagnosis (CAD) relies on accurate segmentation of healthcare images. When it comes to glaucoma, diabetes-related edema, and age-related macular degeneration, correct segmentation of retina images can assist detect these conditions. Medical image segmentation has been greatly improved by deep convolutional neural networks (CNNs) recently. The problem is that all approaches presume that the images in the datasets are drawn over similar set of images. There are several scanner suppliers in clinical practice, which results in domain changes due to differences in the way the images are seen, perceived, and perceived quality [1]. The recognition efficiency with a deep learning model on a fresh image collection with a big decentralized inconsistency gradually degrades as a result of these domain changes. Domain adaption (DA) and domain generalisation (DG) strategies have been developed to reduce the distribution incompatibility across domains to relieve the problem of domain shift. With enough labelled data from a relevant but unrelated source domain, DA seeks to enhance its performance in its target domain [2]. Distribution alignment and style transfer are common techniques used by DA systems to learn domain-invariant representations. Nevertheless, it still demands pre-unlabeled images from the target domain for training. A lack of generalizability is one of the possible drawbacks of DA approaches. Like DA, DG seeks to build a model that can be applied to a wide range of unrelated areas. Researchers have looked into using DG for medical image analysis because of its impressive results in image analysis. For image-feature augmentation, the researchers developed an extra division for acquiring data from many realms. These DG approaches have a key limitation in that they require information from previous or more representations for training. Large volumes of medicinal metaphors after many fields are not obtainable in experimental practice in this scenario [3]. Figure 1: Sample retinal images from the dataset A Novel Medical Image Segmentation Model with Domain Generalization Approach R Gomathi 1 , S Selvakumaran 2 1 Department of ECE, University College of Engineering Dindigul, Tamilnadu, India, gomathiaudece@gmail.com 2 Department of EEE, PSNA CET, Dindigul, Tamilnadu, India, selvakumaran1977@gmail.com *Correspondence: R Gomathi; Email: gomathiaudece@gmail.com ARTICLE INFORMATION Author(s): R Gomathi, S Selvankumaran; Special Issue Editor: Dr. S. Gopalakrishnan ; Received: 18/04/2022; Accepted: 03/06/2022; Published: 30/06/2022; E- ISSN: 2347-470X ; Paper Id: 0422SI-IJEER-2022-16; Citation: 10.37391/IJEER.100242 Webpage-link: https://ijeer.forexjournal.co.in/archive/volume-10/ijeer-100242.html This article belongs to the Special Issue on Intervention of Electrical, Electronics & Communication Engineering in Sustainable Development Publisher’s Note: FOREX Publication stays neutral with regard to jurisdictional claims in Published maps and institutional affiliations.