Abstract
Diabetes and heart disease are common chronic diseases, and a lot of people are suffering from them. Diabetes is caused by insulin secretion disorder. Sometimes it causes death in some complications of heart patients. The current chapter is intended to track diabetes and heart patient conditions, monitoring, and classification. The chapter is designed to predict the two mentioned diseases by diagnosing the patient's previous medical history. An electrocardiogram sensor is used to collect data for heart disease and a continuous glucose monitoring sensor is used to collect data for diabetes disease. A band pass filter in the range between 5-15 Hz is used to filter data, artifact removal is done by stationary wavelet transform, independent component analysis is used for feature extraction.
TopIntroduction
A large group of people are going through diabetes and heart diseases. Heart disease is the greatest vital health issue in the 21st century (Khayyam et. al,2013), annual death from heart disease rises from 2.26 million to 4.77 million in the year of 1990 to 2020 (Huffman et. al., 2010). Diabetes mellitus stands as a pervasive health challenge globally, affecting approximately 537 million adults worldwide, with projections estimating a rise to 784 million by 2045. Alongside its metabolic disruptions, diabetes significantly elevates the risk of cardiovascular diseases (CVDs), marking a significant concern for public health systems and individual well-being. Among the cardiovascular complications associated with diabetes, the heightened susceptibility to heart attacks, medically termed myocardial infarctions (MI), stands out as a critical focal point due to its potential for severe morbidity and mortality. Traditional risk assessment models, such as the Framingham Risk Score, have played crucial roles in identifying cardiovascular risk factors. However, these models often provide generalized estimations and may not sufficiently capture the nuanced, individualized risks faced by diabetic patients. Moreover, the dynamic nature of diabetes and its interplay with various physiological and lifestyle factors necessitates a more sophisticated and adaptable approach to risk prediction.
The advent of mobile health (mHealth) technologies, coupled with advancements in machine learning (ML) algorithms, has opened new avenues for personalized healthcare and proactive disease management. The ubiquitous presence of smartphones and wearable devices equipped with sensors capable of monitoring vital signs, physical activity, glucose levels, and other relevant health metrics presents a unique opportunity to gather real-time, high-resolution data. Harnessing this wealth of information through ML-driven predictive modeling holds immense potential for early detection, risk stratification, and targeted interventions in diabetic patients at risk of heart attack. The diabetes cases in India is in 69.1 million people which takes place in 2nd all over world and 1st one is China (Tripathy et. al., 2017). Hence, early diabetic and heart disease prediction and preventive action is basically essential for every single patient. If in development prediction is exactly possible, metabolic ailment diabetes risk issues can be controlled. Large amounts of healthcare data are influential learning tools that can help healthcare experts to analyze and predict diabetes and heart disease. Machine learning based algorithms have developed and result of its ascent. Machine learning approaches are useful in the analyses and predict diabetes disease, presenting a reasonable level of efficiency. Different types of machine learning techniques including SVM and BPNN (Das et. al., 2020) techniques are used to classify the whole dataset in two categories- disease detection or healthy person. The paper also has a clear idea on how the system will be used for disease prediction and rehabilitation (Das et. al., 2022; Ghosh,2022, Bhowmick et. al., 2023; Mazumdar et. al., 2023).
Key Terms in this Chapter
EHRs (Electronic Health Records): EHRs are digital versions of patients' paper charts, providing real-time access to comprehensive patient data for healthcare providers.
LDA (Linear Discriminant Analysis): LDA is a statistical method used for pattern recognition and classification, often applied in machine learning for dimensionality reduction.
AI (Artificial Intelligence): AI refers to the simulation of human intelligence in machines, enabling them to perform tasks such as learning, reasoning, and problem-solving.
IoT (Internet of Things): IoT connects everyday devices to the internet, enabling them to collect and exchange data, which can be used for monitoring and control in various applications.
SWT (Stationary Wavelet Transform): SWT is a signal processing technique used to decompose a signal into different frequency components, often used for noise reduction and feature extraction.
CGM (Continuous Glucose Monitoring): CGM systems continuously measure glucose levels in real-time, helping individuals with diabetes manage their blood sugar more effectively.
BPNN (Backpropagation Neural Network): BPNN is a type of artificial neural network that uses backpropagation for training, allowing it to learn from errors and improve performance over time.
SVM (Support Vector Machine): SVM is a supervised learning algorithm used for classification and regression tasks, which finds the optimal hyperplane to separate different classes in the data.
mHealth (mobile health): Mobile health involves the use of mobile devices and apps to support medical and public health practices, enhancing patient care and health education.
ECG (Electrocardiogram): An ECG is a test that measures the electrical activity of the heart, helping to diagnose various heart conditions by tracking heartbeats.
HIE (Health Information Exchange): HIE refers to the secure sharing of health-related information among different healthcare organizations, aimed at improving the quality and efficiency of care.
KNN (K-Nearest Neighbors): KNN is a simple, non-parametric algorithm used for classification and regression, which predicts the outcome based on the closest data points in the feature space.
VR (Virtual Reality): VR is a technology that creates a simulated environment using computer-generated imagery, allowing users to immerse themselves in a 3D virtual world.
AR (Augmented Reality): AR overlays digital information, such as images or data, onto the real world through devices like smartphones or AR glasses, enhancing the user's perception of reality.