Heart energy signature spectrogram for cardiovascular diagnosis

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Accurate characterization of human heart sounds and murmurs is crucial for diagnosing cardiac conditions.
• Existing spectrogram methods often lack the resolution needed for detailed analysis of complex acoustic signals.

Purpose of the Study:

• To introduce a new method for characterizing the intensity and pitch of human heart sounds and murmurs.
• To establish a visual map of heart sound energy using advanced time-frequency transformations.

Main Methods:

• Utilized recorded heart sounds from a specialized library.
• Employed Wigner-Ville joint time-frequency transformations for signal representation.
• Developed a methodology to separate acoustic contributions of cardiac events by pitch, time, and energy.

Main Results:

• Created a visual map of heart sound energy for both normal and abnormal recordings.
• Demonstrated superior resolution accuracy compared to existing spectrogram methods.
• Presented the characteristic energy signature of an innocent heart murmur with an S3 sound, enabling clear detection of S1, S2, S3 sounds, S2 split, and systolic murmur intensity.

Conclusions:

• The proposed methodology allows for the full quantitative characterization of heart sounds and murmurs.
• Achieved high accuracy in both time and pitch resolution.
• Generated visual images with self-referencing quality for immediate identification of cardiac acoustic features and their variations.