Diagnosis of left ventricular hypertrophy using convolutional neural network.

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Left ventricular posterior wall thickness (LVPWT) is crucial for cardiac diagnosis.
• Current manual measurement methods are subjective and prone to errors.
• Accurate LVPWT measurement is essential for clinical decision-making.

Purpose of the Study:

• To develop an automated system for accurate LVPWT measurement.
• To reduce subjectivity and errors associated with manual echocardiographic measurements.
• To improve the reliability of LVPWT assessment in clinical practice.

Main Methods:

• A convolutional neural network model was developed using TensorFlow for LVPWT positioning.
• Image processing techniques including non-local mean filtering and opening operations were applied.
• Edge detection algorithms, incorporating threshold segmentation and the OTSU algorithm, were used for contour extraction and measurement.

Main Results:

• The proposed method effectively extracts the left ventricular posterior wall contour.
• Accurate thickness measurements were achieved with a relative error less than 15% compared to hospital values.
• The measurement repeatability error was below the clinical acceptable limit of 20%.

Conclusions:

• The automated measurement method offers reduced manual intervention.
• The image processing and measurement techniques are robust and practical.
• This approach holds significant value for clinical applications in echocardiography.