Skip to navigation Skip to main content Skip to footer

Approved Research

Artificial intelligence research for cardiac electrical biomarker identification

Principal Investigator: Dr Joon-myoun Kwon
Approved Research ID: 89372
Approval date: September 30th 2022

Lay summary


Analysis of changes in electrocardiogram will enable the early diagnosis and prevention of various diseases. Several lines of evidence support the use of ECG(Electrocardiogram) data to find association with critical diseases. However, studies that have aimed to predict diseases through changes in electrocardiogram with strong accuracy and state-of-the-art techniques is lacking. We aim to investigate the usefulness of electrocardiogram, such as rhythm and patterns to evaluate their relations with our target outcomes, and to identify a novel cardiac electrical biomarker that can be used to predict changes in patients.

Scientific rationale:

According to a recent study, AI(Artificial Intelligence)-enabled ECGs can be used to detect heart-related diseases and events.

Currently, the relationship between electrocardiogram and disease is a mostly extracted under human knowledge based on conventional medical evidence. However, there is a possibility of finding novel electrophysiological biomarkers by analyzing bigdata using artificial intelligence methods.

Artificial Intelligence analysis methods can analyze waveform data itself, such as an electrocardiogram. It has advantage to explore the relation between the electrocardiogram and the disease through automated feature extraction (a machine learning method that does not require manually curated features).

Project duration: about 3 years

      -       6 months: data preparation and build a database

      -       3~4 months: data pre-processing and evaluation of existing association

      -       1 year: discovery of a novel cardiac electrical biomarker using the deep learning model

      -       6 months: subgroup analysis

      -       3~4 months: summary and supplement of research results

Expected impact on the public health:

Our study will contribute a novel cardiac electrical biomarker that can be used to predict changes in electrocardiogram. The marker can be used for the early detection and prevention of various diseases. Specifically, the cardiac electrical biomarkers can be used to monitor high-risk patients with frequent diseases non-invasively and continuously using mobile devices. Furthermore, an intervention strategy can be established to prevent or delay the progression of diseases.