Skip to content
Conference Open access

Deep Learning-based O-SBGC-LSTM Framework for Early Diabetes Risk Detection

Sep 2026 · 2026 10th International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC) · pp. 1804-1812 · 0 citations · 20 references

TL;DR

The model proposed incorporates an intelligent preventive module based on a fuzzy inference system for giving intelligent recommendations in the field of healthcare and achieves higher reported accuracy, precision, recall, F1-score, than the compared traditional machine learning models and other deep learning based models.

Abstract

Diabetes mellitus is a chronic metabolic disease that is becoming more common worldwide. Patients have an increased risk for cardiovascular problems, particularly problems associated with Coronary Heart Disease (CHD). One of the most effective ways to lessen mortality rates for patients suffering from diabetes and to enhance their quality of care is to identify diabetes-related complications sooner. While traditional machine learning and modern methods of deep learning could be beneficial, they might not have the ability to learn and predict efficiently from clinical data with rich spatial and temporal features. In this context, this work uses the Eurygaster Optimization Algorithm (EOA) to tune the hyperparameters of the Optimal Scrutiny Boosted Graph Convolutional Long Short Term Memory (O-SBGC-LSTM) for better prediction capability and accuracy to detect early signs of diabetes. The proposed framework healthcare data through data preprocessing, graph-based feature extraction, sequence learning, and deep learning and hyperparameter optimization. In EOA, there is a consideration of the hyperparameters for learning rate, batch size, number of hidden units and dropout rate, in which optimization is done to minimize overfitting and enhance convergence. The model proposed incorporates an intelligent preventive module based on a fuzzy inference system for giving intelligent recommendations in the field of healthcare. The experimental results shows that the EOA-optimized O-SBGC-LSTM model achieves higher reported accuracy, precision, recall, F1-score, than the compared traditional machine learning models and other deep learning based models. The proposed framework is useful for early detection of diabetes and prevent cardiovascular complications.

Read PDF

Similar papers

Conference Open access 2026

Cardiovascular Disease Prediction Using Hybrid CNN-LSTM Architecture in Deep Learning

A hybrid deep learning framework is proposed, which learns discriminative clinical representations and health patterns over time together with Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks.

M. M., A. S, K. B · 0 citations
Conference Open access Sep 2026

Adaptive Feature Dependent Deep Transfer Learning Model for Improved Disease Prediction Using Big Data

The use of big data and deep learning methods in healthcare analytics is becoming more important in terms of effective disease prediction in an early stage. Nevertheless, the current methods are prone to low performance because of incomplete, heterogeneous, and inconsistent clinical data that are gathered through vario...

P. Malathi, R. Parameswari · 0 citations
Open access 2026

Deep Learning-Based Multimodal Fusion for Early Prediction of Cardiovascular Risk Using Clinical and Imaging Data

This work proposes a multimodal deep fusion framework with attention for high accurate cardiovascular risk stratification using the integration of medical images and clinical data and demonstrates that this adaptive fusion strategy outperforms simple concatenation baselines.

Amit Thakur, Sarita Kumari, Sarita Thakur et al. · 0 citations
Open access Oct 2026

CNN-BiLSTM based model for heart disease classification using deep learning techniques

Heart disease remains a main cause of death worldwide, showing the necessity for accurate and timely automated diagnostic systems. Hence, precise and early detection is necessary and is done by some existing machine learning (ML) and deep learning (DL) algorithms. Existing ML and DL models, like convolutional neural ne...

M. Azhar, L. Gladence · 0 citations
Open access Sep 2026

A multimodal risk stratification model using deep learning and tabular data

A deep learning–based framework for the simultaneous prediction of CVD and stroke risks using tabular health data and the potential of interpretable deep learning models to support early, data-driven risk stratification of cardiovascular and stroke risks directly from cross-sectional tabular data is proposed.

Abdelrahman Alaa Sadik, M. Morsey, T. Nazmy et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.