Automatic left ventricle segmentation using iterative thresholding and an active contour model with adaptation on...

Area of Science:

• Cardiology
• Medical Imaging
• Image Processing

Background:

• Accurate quantification of cardiac output and myocardial mass is crucial for clinical practice.
• Existing methods for left ventricle (LV) segmentation have limitations in precision and efficiency.

Purpose of the Study:

• To introduce and validate the iterative thresholding and active contour model with adaptation (ITHACA) algorithm for automatic LV segmentation.
• To assess the accuracy and clinical utility of ITHACA compared to manual tracing and commercial software.

Main Methods:

• The ITHACA algorithm employs region growing with iterative thresholding for endocardial segmentation.
• Epicardial segmentation is achieved using an active contour model guided by endocardial borders and estimated myocardial signals.
• The algorithm was evaluated in 38 patients against manual tracing and MASS Analysis software.

Main Results:

• ITHACA demonstrated significant improvements over MASS software in defining myocardial borders.
• ITHACA showed good agreement with manual tracing for blood volume and myocardial mass quantification.
• The differences observed with ITHACA were smaller than those between manual tracing and MASS software.

Conclusions:

• The ITHACA algorithm provides accurate and reliable automatic left ventricle segmentation.
• ITHACA offers a valuable tool for clinical practice, enhancing the quantification of cardiac parameters.
• The proposed method represents a substantial advancement in automated cardiac image analysis.