ABNORMAL HEART ARRYTHMIA DETECTIONS AND CLASSIFICATIONS FROM ELECTROCARDIOGRAM (ECG) USING DEEP LEARNING

In this project we will use the following steps: Data preprocessing, Transformation of electrocardiogram signal to spectrum, Abnormality detection of heart, and arrhythmia classification. In this project, we will use an ECG acquisition sensor, a mobile application, and a Holter monitoring device for

Project Title

ABNORMAL HEART ARRYTHMIA DETECTIONS AND CLASSIFICATIONS FROM ELECTROCARDIOGRAM (ECG) USING DEEP LEARNING

Project Area of Specialization

Artificial Intelligence

Project Summary

In this project we will use the following steps: Data preprocessing, Transformation of electrocardiogram signal to spectrum, Abnormality detection of heart, and arrhythmia classification. In this project, we will use an ECG acquisition sensor, a mobile application, and a Holter monitoring device for data collection. ECG acquisition sensor is a Certified medical device that is available in the market. ECG acquisition sensor will use a strap or belt to attach to the skin, it is a small device and easy to attach and detach, it is waterresistant, it can easily use for weeks without interference in daily activities. the ECG acquisition sensor device has a Bluetooth connection and data transfer to the mobile application. Mobile application that records the data from the ECG acquisition sensor. It will record data for a long time as the patient wish or doctor requirement and will send the data to the database instantly and the program adds to the mobile application that runs on the data for every 5 seconds and show the Rhythm type on the screen of the normal App or arrhythmia. After recording the monitoring will first manually and then it will be done by the model later. On running the model create a report like the Holter machine. The first 24 hours report of our model will compare to the Holter report. the total recording of our device that will be more than 24 hours will be compared with the recording of the Holter device of 24 hours and will create a report automatically

Project Objectives

Our main objectives will be: • To make the device cheaper as compared to previously developed devices. • Due to the lack of ECG classification experts, it is impossible to diagnose the correct disease so the device automatically detects the correct disease. • To develop a low-cost ECG classification algorithm which will be made possible due to Machine Learning and Deep Learning. • The device automatically detects the abnormal activity of the heart using an ECG signal. • The device will be easily used for Heart patients. • It takes less time to give the result is compared to manual. • It will replace the position of a cardiologist. • It will user friendly it means it can be used anywhere. • The device will be available in the country due to domestic initiatives.

Project Implementation Method

Frist of all we will study relevant literature and will learn about the signal of ECG and different arrhythmias. we will study different result of previous work on ECG papers. In this project we will use an ECG acquisition senser, a mobile application and Holter monitoring device for the data collection. ECG acquisition senser will be a Certified medical device that is available in the market. ECG acquisition senser will use a strap or belt to attached to the skin, it will be a small device and easy to attached and detached, it will be water resistant, it can be easily use for weeks without interference in the daily activities. The ECG acquisition senser device have Bluetooth connection and the data transfer to the mobile application. That mobile application will record the data from the ECG acquisition senser. It will records data for long time as the patient wish or doctor requirement and will send the data to the database instantly and the program add to the mobile application that run on the data for every 5 second and show the Rhythm type on the screen of the App that is normal or arrythmia. After recording the monitoring will first manually and then it will be done by model later. On running the model create a report like Holter machine. The first 24 hours report of our model will compare to the Holter report and then total recording of our device that will more than 24 hours will be compare with the recording of Holter device of 24 hours and will create a report automatically. The methodology planned to be used to construct to tackle the problems efficiently.

DATASET

DATA PREPROCESSING

PROPOSED MODEL

TRAINING AND TESTING THE MODEL

FINAL BUILD

Benefits of the Project

The benefits for this project is the service of humanity by introducing a Vision Aid Device which will be cheaper and be designed for marginalized class of society as compared to all such devices marketed today. This would be a great service to humanity on one side and would be the cheapest facility for the poor patients, on the other. The device would become a sense of living for those who live in darkness.

Technical Details of Final Deliverable

We will provide hardware and software for the Arrhythmia detection model. Make an App or JUI for the model and we will get an ECG device that is medically available for the collection of the ECG dataset and the app will detect the Arrhythmia in an instant and will give result.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Medical

Other Industries

IT

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT)

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

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