Automated 4-dimensional regional myocardial strain evaluation using cardiac computed tomography

Area of Science:

• Cardiology
• Medical Imaging
• Biomedical Engineering

Background:

• Myocardial regional function evaluation is typically subjective ('eyeballing').
• A need exists for objective, repeatable measures of myocardial performance.
• Cardiac computed tomography (CT) offers detailed anatomical data but requires robust functional analysis tools.

Purpose of the Study:

• To evaluate the clinical utility of novel regional myocardial strain software using cardiac CT datasets.
• To assess the software's ability to provide objective and quantifiable regional myocardial function analysis.
• To compare the software's results against visual assessment and echocardiography.

Main Methods:

• Retrospectively gated cardiac CT datasets from 93 patients were analyzed.
• Novel software employing a finite element-based tracking algorithm calculated circumferential (CS), longitudinal (LS), and radial (RS) strains.
• Results were compared to cardiologist visual assessment and echocardiography speckle strain in a subset of cases.

Main Results:

• The automated software successfully analyzed 93/106 cases with minimal interaction.
• Peak CS, LS, and RS strains effectively differentiated between normal, hypokinetic, and akinetic myocardial segments.
• ROC analysis showed good differentiation capabilities for CS (AUC 0.84) and RS (AUC 0.80), with moderate agreement with echocardiography.

Conclusions:

• Automated 4D regional strain analysis of CT datasets provides an objective, quantifiable map of myocardial regional function.
• The software demonstrates good correspondence with visual analysis and differentiates effectively between normal and abnormal segments.
• This technology offers a promising advancement for assessing myocardial performance.