Development of IoT enabled wearable system for pre-heart failure detection

New technologies have been commonly adopted, paving the way for modern healthcare systems. There has been a lot of progress in medical science but in addition, computer technology and interactive systems produce various services that are far closer to the patients in question. The availability of we

Project Title

Development of IoT enabled wearable system for pre-heart failure detection

Project Area of Specialization

Internet of Things

Project Summary

New technologies have been commonly adopted, paving the way for modern healthcare systems. There has been a lot of progress in medical science but in addition, computer technology and interactive systems produce various services that are far closer to the patients in question. The availability of wearable devices helps measure vital medical parameters, such as temperature, blood pressure, heart rate, oxygen level concentration in blood and cardiovascular activities of the human heart, etc. These devices can also update the physicians as well as patients about their medical condition through smart internet communication with their mobile phones. The purpose of this device is to integrate medical sensors e.g. temperature sensor, heart rate sensor and ECG sensor to the microcontroller with built-in WIFI. Sensors’ data is subjected to analyze through microcontroller and further algorithms are applied to it to extract useful information about human health from this raw data. Based on the QRS complexity of the Electrocardiogram and irregularity of different intervals in ECG curves, algorithms are subjected to implement pre-heart attack detection. Fast and irregular heartbeats known as palpitations are major symptoms of heart attacks. Furthermore, checks are applied to detect abnormalities in human health. This data is further communicated over WIFI to an online database where all the useful information related to body temperature, heart rate, SpO2 concentration and ECG of a patient can be stored over time. This database will be accessed by the physician as well as guardians of the patient using their mobile phone. They can access this informational data any time using their mobile phone by connecting to an online server through WIFI.

Such devices are being manufactured by some companies and being used in private but many expensive hospitals where an ordinary person can’t afford his treatment. One of our purposes is to make such a device with very low cost so that it can be given in access to ordinary home users to help them improve their health standards as approved by world health organization. Akram Medical Complex Lahore is our medical industry partner in the implementation, clinical trials, and commercialization of this project.

Project Objectives

The primary goal of this project is to demonstrate an IoT system using multiple sensors working synchronously to collect data of different physical movements i.e. body temperature, heart rate, and ECG waveform. With the help of Wi-Fi enabled embedded controller, the above-mentioned parameters are collected and displayed on the android based application. The scope of this project includes various techniques such as signal processing and machine learning algorithms to predict sudden cardiac arrest. This project is basically a health monitoring system via basic parameters with the value addition of pre-heart failure prediction.

1. To design a low-cost IoT wearable device to predict the cardiac arrest.
2. Design a complete health monitoring system based on vital signs of the body.
3. To create the feasibility of a detailed clinical study for pre heart failure detection using our proposed algorithm based on the complexity of PQRST intervals in the ECG waveform
4. To commercialize the IoT device on a large scale for easy use in households and hospitals.

Project Implementation Method

We provide a predictive system for smartphones that can warn users about their irregular ECG waveform patterns. As the ECG rhythm becomes erratic which may induce myocardial infarction, it uses the detection of an irregular pulse activity to alert consumers to possible heart failure. The system proposed is beneficial not only for elderly people but also to assess cardiac illness for infants, adults, and patients with strokes.

The basic features of the heart are extracted using an ECG sensor i.e. the Heart signal. Three electrodes are attached to the chest of the test subject which senses the heartbeats as voltage inputs. These graphs can be plotted on the MATLAB to further process the data. From the ECG signal, the basic features are the P, Q, R, S, T peaks which can be classified as the different working segments of the heart in which heart receives the blood from the body and pumps the blood to the different parts of the body

From these features of the ECG signal, we can derive a number of conclusions by simply evaluating the position and magnitude of each of the peaks present in the ECG signal. One of the very basic features is the heart rate. It is actually the number of heartbeats in a minute. This can be calculated through ECG when we extract the time duration between two consecutive R intervals and evaluate them in a minute, we will get the heart rate.

The heart rate can be used to identify two of the diseases named Tachycardia and Bradycardia. Tachycardia means when your heart rate goes above 100 beats per minute and Bradycardia means that the heart rate falls below 60 beats per minute.

The prediction of a heart attack requires a few numbers of parameters that should be detected to successfully predict heart failure. After doing research of a few parameters we finalized that the following parameters will be detected using the sensors on which we will evaluate our algorithms:

•          Body Temperature

•          Heart rate

•          ECG

These three parameters will give different levels of heart attack predictions. For example, the abnormal body temperature will be level 1 danger towards heart attack and abnormal heart rate will be level 2.

Our algorithm is based on an analysis machine-learning predictive model known as the Decision Tree. This model chooses a new example's goal value based on many

characteristic data values. This model is used in our algorithm to indicate whether the patient

has an acute heart arrest or not and to include available data in the extracted function.

We developed a unique algorithm using a regular statistical deviation study of the decision tree model. Now we analyze the mechanism by which derived functions are interpreted at the decision tree. We determine whether or not the functions are abnormal with a standard statistical deviation analysis. The decision tree layout corresponding to the prediction algorithm can be seen in Figure 3. Our algorithm uses 0 to 4 alerting thresholds to assess the degree of abnormality for every slot.

Benefits of the Project

In a developing country like Pakistan, there are a lot of opportunities which meant to solve the relevant problems that the health industry is facing. There are only 1219 public sector hospitals for a population of 200 million people. It creates a high deficiency of doctors as well as nursing staff. We are facing the severe problems of having only one doctor for 1000 patients and one nursing staff for 50 patients. These figures are highly alarming for health quality insurance to the people of Pakistan which is the basic right of every human being. So, there needs to be some technological revolution that can manage the deficiency of human staff by balancing the workload.

These health monitoring devices reduce the need of nurses by automatically measuring the vital signs of the patient (wearing the device) as well as decreases the burden on doctors by sending patient’s data to the display screen in the doctor’s room. These devices have introduced a remarkable trend in the adoption of technological advancement in health monitoring techniques enhancing public satisfaction by providing quality insurance to their health. While 82% of users feel that these devices have enhanced their lifestyle. 88% of physicians want their patients to monitor their vital signs at home so that they could be informed and updated by their health without visiting the doctor.

Technical Details of Final Deliverable

The main objective of this project is to build a complete prototype of IoT enabled wearable system including

• Complete Sensors calibration and their compatibility with the microcontroller.
• Microcontroller integration with MATLAB to process the ECG waveform to extract the P-Q-R-S-T intervals.
• Hardware and web/android app integration using the cloud services to monitor the vital signs of the body.
• Efficient machine learning-based algorithm for the prediction of pre heart failure using decision trees.

Final Deliverable of the Project

Hardware System

Core Industry

Medical

Other Industries

Core Technology

Internet of Things (IoT)

Other Technologies

Sustainable Development Goals

Good Health and Well-Being for People, Decent Work and Economic Growth, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com