Fusion-ADiNet - a multi-level framework for enhanced diabetes and Alzheimer’s disease detection using chimp-whale fusion estimation

Sep 30, 2026

Fusion-ADiNet presents a multi-level fusion framework aimed at improving diabetes (DM) and Alzheimer’s disease (AD) detection. The method addresses limitations of prior approaches caused by low diagnostic accuracy, high computational complexity, and difficulty modeling subtle cross-disease correlations in high-dimensional clinical data. It introduces the Chimp-Whale Fusion Estimator (CWFE) optimization algorithm and is designed for flexible extension to other diseases or datasets, supporting more effective healthcare analytics and timely clinical decision-making.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages21

Reading Time53 min

AccessFree to download

Background and Motivation

Limitations of Existing Disease Detection Methods

Proposed Method: Fusion-ADiNet

Multi-level Fusion Framework

Chimp-Whale Fusion Estimator (CWFE)

Diabetes Mellitus Overview

Hyperglycemia and Insulin Dysfunction

Complications and Subtypes

Type 1 Diabetes (T1DM)

Type 2 Diabetes (T2DM)

Gestational Diabetes (GDM)

[www. nature.com/scientificreports](www. nature.com/scientificreports) OPEN Fusion-ADiNet: a multi-level framework for enhanced diabetes and Alzheimer’s disease detection using chimp-whale fusion estimation Roobini MS1,5􀀍, Mong-Fong Horng2, Siva Shankar S3 & NagarajanG4 Rising cases of diabetes andAD are the two biggest health-related issues world-wide; this greatly hampers the quality of life among the afflicted persons. Early identification of diabetes most especially correlated with neurodegenerative conditions such as Alzheimer’s guarantees early intervention and hence proper management. However, the existing approaches for disease detection have been suffering from a series of limitations in poor diagnostic accuracy, high computational complexity, and the lack of an effective model that could properly handle the intricate correlations between diabetes and AD. Most of the existing methods for diabetes and AD detection rely on traditional machinelearning algorithms or heuristic optimization approaches, which are not capable of handling high dimensionality and complex clinical data. The models also find it extremely difficult to represent the subtle relationship between the two diseases, leading to unsatisfying performance in practical applications. It is therefore imminent to develop much more advanced methodologies with integration that could improve accuracy in prediction, enabling better decision-making in clinics. To overcome the limitations of existing methods, in this paper, we propose a new approach called Fusion-Alzheimer’s Diabetes Network (Fusion-ADiNet) by introducing a multi-level fusion framework for disease detection. The main novelty in this method lies in the newly designed Chimp-Whale Fusion Estimator (CWFE) optimization algorithm. Furthermore, the Fusion-ADiNet framework is quite flexible, so extending it to other diseases or datasets in the future will be very easy. This work contributes much to the field of healthcare analytics and opens new perspectives toward more effective diagnostic tools for timely detection of diabetes and Alzheimer’s disease. Keywords Diabetes mellitus (DM), Alzheimer diseases (AD), Deep learning, Disease diagnosis, Classification, Optimization Diabetes mellitus (DM) is a multi-factorial and long-term disease process, which is mainly clinically expressed as hyperglycemia, a state of elevated blood glucose levels. This state results from derangements in insulin release, insulin sensitivity, or both 1,2. It is a hormone called insulin that is synthesized by the beta cells found in the pancreas to help cells move sugar from glucose into the cell for it to produce energy or glucose to be stored. When regulating mechanisms that control insulin production and usage are thrown off, then one experiences high blood sugar levels and if not controlled, cause serious harm for many body organs and tissues. Diabetes mellitus chronic hyper-glycemia is the primary cause of complications that target important organs of the body includingthe heart, kidneys, eyes, nerves, and blood vessels. These complications can in turn cause cardiovascular diseases, diabetic nephropathy, retinopathy, neuropathy, and peripheral vascular diseases which strongly affect the quality of life and even contributes to mortality. Diabetes mellitus is categorized into three primary subtypes based on its underlying causes and clinical manifestations: two primary forms that include type 1 diabetes (T1DM), 1National KaohsiungUniversity of Science andTechnology, No. 415, Jiangong Road, Kaohsiung,Taiwan. 2Department of Electronics Engineering, National Kaohsiung University of Science and Technology, No. 415, Jiangong Road, Kaohsiung, Taiwan. 3Department ofCSE, KG Reddy College of Engineering and Technology, Hyderabad, Telangana, India. 4Department of Computer Science and Engineering, Sathyabama Institute of Science and Technology, Chennai, India. 5Sathyabama Institute of Science and Technology, Chennai, India. 􀀍 em

FAQs

What problem does Fusion-ADiNet target in diabetes and Alzheimer’s detection?

It targets low diagnostic accuracy, high computational complexity, and inadequate modeling of intricate correlations between diabetes and Alzheimer’s using high-dimensional clinical data.

What is the key novelty of the Fusion-ADiNet framework?

The work proposes a new Chimp-Whale Fusion Estimator (CWFE) optimization algorithm within a multi-level fusion framework for disease detection.

How can Fusion-ADiNet be applied beyond diabetes and Alzheimer’s?

The framework is described as flexible, making it easy to extend to other diseases or datasets in future work.

Sep 30, 2026

Uploaded by Sophia Brooks