Semia: semi-automatic interactive graphic editing tool to annotate ambulatory ECG records

Area of Science:

• Cardiology
• Medical Informatics
• Biomedical Engineering

Background:

• Accurate annotation of ST segment events in ambulatory electrocardiogram (ECG) records is crucial for diagnosing cardiac conditions.
• Existing methods for analyzing long-term ECG data can be time-consuming and labor-intensive.
• The Long-Term ST database (LTST DB) requires efficient annotation tools for its multinational dataset.

Purpose of the Study:

• To design and develop a semi-automatic interactive graphic editing tool named Semia.
• To facilitate the accurate annotation of transient ischemic ST segment episodes and other ST segment events in 24-hour ambulatory ECG records.
• To enhance the efficiency and usability of ECG data annotation for expert annotators.

Main Methods:

• Development of a specialized interactive graphic editing tool (Semia) with semi-automatic annotation capabilities.
• Implementation of features for data representation, multi-resolution viewing, and interaction.
• Inclusion of manual adjustment of heart-beat fiducial points and both manual and automatic editing of annotations.
• Integration of efficient waveform display, feature-vector time-series visualization, and dynamic interface controls.

Main Results:

• The Semia tool demonstrated efficiency, user-friendliness, and usability through its design and automated procedures.
• Human expert annotators successfully utilized Semia to annotate the Long-Term ST database (LTST DB).
• The tool supported paperless annotation editing across geographically dispersed sites, enhancing collaborative research.

Conclusions:

• The Semia tool provides an effective solution for annotating ST segment events in ambulatory ECG data.
• Its design facilitates efficient, user-friendly, and accurate analysis of complex cardiac rhythm data.
• Semia supports collaborative, multi-site research by enabling paperless annotation workflows.