Cardio Predict IoT (CPI): Real Time Heart Disease Prediction Using Cloud Enhanced Machine Learning

Author: Antim Dev Mishra, Bindu Thakral, Alpana Jijja and Nitin Sharma

Download PDF

Abstract

Cardio Predict IoT (CPI) is an innovative system designed for real-time heart disease prediction utilizing cloud-enhanced machine learning. This study leverages a comprehensive dataset of real-time vital signs collected from a large cohort, including body temperature, pulse rate, systolic and diastolic blood pressure, and oxygen saturation. The integration of IoT sensors with feature-based advanced machine learning algorithms demonstrates superior performance compared to other state-of-the-art techniques. The research methodology encompassed data collection from IoT sensors, preprocessing, and the application of various machine learning algorithms for heart disease prediction. Notably, the multilayer perceptron model exhibited exceptional performance, achieving the highest accuracy of 97.28% and an area under the curve (AUC) of 0.95 when cross-validation was applied. This study highlights the significant potential of combining cloud-based machine learning and IoT integration in predictive healthcare. The CPI system offers a scalable and responsive solution for proactive heart disease management, potentially revolutionizing early detection and prevention strategies in cardiology. The findings underscore the importance of real-time data analysis in healthcare and demonstrate the feasibility of using IoT devices for continuous patient monitoring. By leveraging cloud computing resources, the CPI system can process vast amounts of data rapidly, enabling timely interventions and personalized care plans. The results of this research suggest that the CPI system could play a crucial role in transforming cardiovascular healthcare, offering a promising approach to reducing the global burden of heart disease through early prediction and intervention

Keywords

Realtime healthcare, cardiovascular diseases, heart stroke prediction, and Internet of Things (IoT), ML

Conclusion

The Cardio Predict IoT (CPI) system demonstrates significant advancements in real-time heart disease prediction through the integration of IoT devices and cloud-based machine learning. Our results, particularly the high accuracy (97.28%) and AUC (0.95) achieved by the multilayer perceptron model, align with and even surpass recent developments in the field. The superiority of our multilayer perceptron model supports the findings of Dobrovska and Nosovets (2021), who developed a classifier based on a multilayer perceptron using genetic algorithms and decision trees. Our model's performance also aligns with the work of Tang (2021), who achieved promising results using logistic regression and random forest models for heart disease prediction. The real-time data collection and analysis capabilities of CPI address a critical gap identified by Aung (2020), who emphasized the need for IoT applications in healthcare for continuous monitoring. Our system's ability to process data from multiple IoT sensors simultaneously aligns with the approach suggested by Kakria et al. (2015) for remote cardiac patient monitoring. The integration of cloud computing in our system allows for scalable and rapid data processing, a crucial factor highlighted by Wankhede et al. (2020) in their comparative study of cloud platforms. This approach enables the timely interventions and personalized care plans that Zullig (2018) identified as key components in reaching individuals with chronic conditions efficiently. Our use of a comprehensive dataset including various vital signs supports the findings of Kaur et al. (2022), who demonstrated that incorporating multiple physiological parameters significantly enhances the accuracy of early stroke prediction methods. The large cohort size in our study addresses the limitations of small sample sizes noted in previous studies, as discussed by Hazra et al. (2017) in their review of heart disease diagnosis and prediction techniques. The high performance of our system in handling real-time data aligns with the work of Li (2022), who emphasized the importance of integrating various data sources and using Google Colab for deep learning modeling in disease prediction. Our approach to data preprocessing, particularly in handling imbalanced data and noise reduction, addresses challenges identified by Ramasamy and Nirmala (2017) in their study on disease prediction using data mining techniques. The potential of CPI to revolutionize early detection and prevention strategies in cardiology is supported by the findings of Sahoo and Jeripothula (2020), who demonstrated the efficacy of machine learning techniques in heart failure prediction. Furthermore, our system's use of IoT and machine learning aligns with the transformative effects on healthcare described by Hussain et al. (2021); Tehseen et al., 2021; Sharmila and Santhosh 2018). The multilayer perceptron's exceptional performance in our study is consistent with recent trends in machine learning applications for heart disease prediction, as noted by Santhana Krishnan et al. (2019); Bhaskaru (2020). The integration of IoT-based models with data mining techniques, as implemented in our CPI system, builds upon the work of Chavan and Sonawane (2017), further enhancing the accuracy and real-time capabilities of heart disease risk prediction. In conclusion, the Cardio Predict IoT system represents a significant step forward in the application of IoT and machine learning technologies to cardiovascular health management. Its high accuracy, real-time capabilities, and scalability position it as a promising tool for improving patient outcomes and reducing the global burden of heart disease, as highlighted by the World Health Organization (2022). The CPI system's innovative approach addresses many of the challenges and limitations identified in previous studies, paving the way for more effective and personalized cardiovascular care

References

-

How to cite this article

Antim Dev Mishra, Bindu Thakral, Alpana Jijja and Nitin Sharma (2023). Cardio Predict IoT (CPI): Real Time Heart Disease Prediction Using Cloud Enhanced Machine Learning. Biological Forum – An International Journal, 15(5a): 773-781.