A ROBUST ARTIFICIAL NEURAL NETWORK APPROACH FOR EARLY  DETECTION OF CARDIAC DISEASES

Ghulam Gilanie; Syed Naseem Abbas; Sana Cheema; Akkasha Latif; Muhammad Ahsan; Hina Shafique; Anum Saher; Syeda Naila Batool; Fatima Bibi; Hunaina Arshad Chohan; Iqra Mubeen

A ROBUST ARTIFICIAL NEURAL NETWORK APPROACH FOR EARLY DETECTION OF CARDIAC DISEASES

Authors

Ghulam Gilanie
Syed Naseem Abbas
Sana Cheema
Akkasha Latif
Muhammad Ahsan
Hina Shafique
Anum Saher
Syeda Naila Batool
Fatima Bibi
Hunaina Arshad Chohan
Iqra Mubeen

Keywords:

Early-stage detection, artificial neural network, cardiac-related disease

Abstract

The research introduces a method focused on the early and precise prediction of cardiac conditions, enhancing treatment effectiveness. It leverages signal processing techniques for feature extraction. Utilized in this study are ten types of metal oxide semiconductor sensors to detect various gases released by the human body. Also, ECG, SPO2, and oxygen sensors contribute to data collection. The study conducts experiments involving groups of 5, 10, 15, and 20 participants, each providing 1000 unique features. These analog signals are converted to digital via Arduino. A specialized model architecture, trained on this newly created dataset, evaluates performance metrics such as sensitivity, f-measures, accuracy, and specificity. This model, designed to identify human odors, achieves an accuracy exceeding 85%.

Downloads

Published

2025-04-16

How to Cite

Ghulam Gilanie, Syed Naseem Abbas, Sana Cheema, Akkasha Latif, Muhammad Ahsan, Hina Shafique, Anum Saher, Syeda Naila Batool, Fatima Bibi, Hunaina Arshad Chohan, & Iqra Mubeen. (2025). A ROBUST ARTIFICIAL NEURAL NETWORK APPROACH FOR EARLY DETECTION OF CARDIAC DISEASES. Spectrum of Engineering Sciences, 3(4), 499–511. Retrieved from https://sesjournal.com/index.php/1/article/view/272