Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems

Varsha Arya; Akshat Gaurav; Priyanka Chaurasia; Kwok Tai Chui; Brij B. Gupta; Ramakrishnan Raman

doi:10.1109/ICCE67443.2026.11449841

Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems

Varsha Arya
, Akshat Gaurav
, Priyanka Chaurasia
, Kwok Tai Chui
, Brij B. Gupta
, Ramakrishnan Raman

Electronic Engineering and Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Diabetes is one of the most prevalent chronic diseases worldwide, requiring timely diagnosis and continuous monitoring to prevent severe health complications. With the growing integration of artificial intelligence in healthcare, smart and efficient models are essential for real-time analysis on consumer health devices. This study proposes a deep learning framework that integrates categorical embeddings with continuous medical features and introduces a novel Delta-Attention mechanism to dynamically capture variations in feature importance during training. Evaluated on the Kaggle diabetes dataset, the model achieved an accuracy of 91.98%, precision of 94.92%, recall of 91.98%, and F1-score of 92.91%. Through ablation and comparative analyses, the model demonstrated superior inter-pretability and computational efficiency, making it highly suitable for on-device diabetes prediction in consumer health monitoring systems.

Original language	English
Title of host publication	2026 IEEE International Conference on Consumer Electronics, ICCE 2026
ISBN (Electronic)	9798331553432
DOIs	https://doi.org/10.1109/ICCE67443.2026.11449841
Publication status	Published - 2026
Event	2026 IEEE International Conference on Consumer Electronics, ICCE 2026 - Dubai, United Arab Emirates Duration: 3 Feb 2026 → 5 Feb 2026

Publication series

Name	Digest of Technical Papers - IEEE International Conference on Consumer Electronics
ISSN (Print)	0747-668X
ISSN (Electronic)	2159-1423

Conference

Conference	2026 IEEE International Conference on Consumer Electronics, ICCE 2026
Country/Territory	United Arab Emirates
City	Dubai
Period	3/02/26 → 5/02/26

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Consumer Health Monitoring
Deep learning
Delta-Attention mechanism
Diabetes prediction

Access to Document

10.1109/ICCE67443.2026.11449841

Cite this

Arya, V., Gaurav, A., Chaurasia, P., Chui, K. T., Gupta, B. B., & Raman, R. (2026). Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems. In 2026 IEEE International Conference on Consumer Electronics, ICCE 2026 (Digest of Technical Papers - IEEE International Conference on Consumer Electronics). https://doi.org/10.1109/ICCE67443.2026.11449841

@inproceedings{8e2b992ff61643efa5f7eeeb1288c2b6,

title = "Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems",

abstract = "Diabetes is one of the most prevalent chronic diseases worldwide, requiring timely diagnosis and continuous monitoring to prevent severe health complications. With the growing integration of artificial intelligence in healthcare, smart and efficient models are essential for real-time analysis on consumer health devices. This study proposes a deep learning framework that integrates categorical embeddings with continuous medical features and introduces a novel Delta-Attention mechanism to dynamically capture variations in feature importance during training. Evaluated on the Kaggle diabetes dataset, the model achieved an accuracy of 91.98\%, precision of 94.92\%, recall of 91.98\%, and F1-score of 92.91\%. Through ablation and comparative analyses, the model demonstrated superior inter-pretability and computational efficiency, making it highly suitable for on-device diabetes prediction in consumer health monitoring systems.",

keywords = "Consumer Health Monitoring, Deep learning, Delta-Attention mechanism, Diabetes prediction",

author = "Varsha Arya and Akshat Gaurav and Priyanka Chaurasia and Chui, \{Kwok Tai\} and Gupta, \{Brij B.\} and Ramakrishnan Raman",

note = "Publisher Copyright: {\textcopyright} 2026 IEEE.; 2026 IEEE International Conference on Consumer Electronics, ICCE 2026 ; Conference date: 03-02-2026 Through 05-02-2026",

year = "2026",

doi = "10.1109/ICCE67443.2026.11449841",

language = "English",

series = "Digest of Technical Papers - IEEE International Conference on Consumer Electronics",

booktitle = "2026 IEEE International Conference on Consumer Electronics, ICCE 2026",

}

Arya, V, Gaurav, A, Chaurasia, P, Chui, KT, Gupta, BB & Raman, R 2026, Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems. in 2026 IEEE International Conference on Consumer Electronics, ICCE 2026. Digest of Technical Papers - IEEE International Conference on Consumer Electronics, 2026 IEEE International Conference on Consumer Electronics, ICCE 2026, Dubai, United Arab Emirates, 3/02/26. https://doi.org/10.1109/ICCE67443.2026.11449841

Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems. / Arya, Varsha; Gaurav, Akshat; Chaurasia, Priyanka et al.
2026 IEEE International Conference on Consumer Electronics, ICCE 2026. 2026. (Digest of Technical Papers - IEEE International Conference on Consumer Electronics).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems

AU - Arya, Varsha

AU - Gaurav, Akshat

AU - Chaurasia, Priyanka

AU - Chui, Kwok Tai

AU - Gupta, Brij B.

AU - Raman, Ramakrishnan

PY - 2026

Y1 - 2026

N2 - Diabetes is one of the most prevalent chronic diseases worldwide, requiring timely diagnosis and continuous monitoring to prevent severe health complications. With the growing integration of artificial intelligence in healthcare, smart and efficient models are essential for real-time analysis on consumer health devices. This study proposes a deep learning framework that integrates categorical embeddings with continuous medical features and introduces a novel Delta-Attention mechanism to dynamically capture variations in feature importance during training. Evaluated on the Kaggle diabetes dataset, the model achieved an accuracy of 91.98%, precision of 94.92%, recall of 91.98%, and F1-score of 92.91%. Through ablation and comparative analyses, the model demonstrated superior inter-pretability and computational efficiency, making it highly suitable for on-device diabetes prediction in consumer health monitoring systems.

AB - Diabetes is one of the most prevalent chronic diseases worldwide, requiring timely diagnosis and continuous monitoring to prevent severe health complications. With the growing integration of artificial intelligence in healthcare, smart and efficient models are essential for real-time analysis on consumer health devices. This study proposes a deep learning framework that integrates categorical embeddings with continuous medical features and introduces a novel Delta-Attention mechanism to dynamically capture variations in feature importance during training. Evaluated on the Kaggle diabetes dataset, the model achieved an accuracy of 91.98%, precision of 94.92%, recall of 91.98%, and F1-score of 92.91%. Through ablation and comparative analyses, the model demonstrated superior inter-pretability and computational efficiency, making it highly suitable for on-device diabetes prediction in consumer health monitoring systems.

KW - Consumer Health Monitoring

KW - Deep learning

KW - Delta-Attention mechanism

KW - Diabetes prediction

UR - https://www.scopus.com/pages/publications/105037423200

U2 - 10.1109/ICCE67443.2026.11449841

DO - 10.1109/ICCE67443.2026.11449841

M3 - Conference contribution

AN - SCOPUS:105037423200

T3 - Digest of Technical Papers - IEEE International Conference on Consumer Electronics

BT - 2026 IEEE International Conference on Consumer Electronics, ICCE 2026

T2 - 2026 IEEE International Conference on Consumer Electronics, ICCE 2026

Y2 - 3 February 2026 through 5 February 2026

ER -

Delta-Attention-Based Deep Learning Model for Diabetes Prediction in Consumer Health Monitoring Systems

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this