Automated speckle tracking algorithm to aid on-axis imaging in echocardiography

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Cardiovascular Diagnostics

Background:

• Echocardiography image acquisition relies on subjective operator skill.
• Standardized views are crucial but difficult to achieve consistently.
• Objective tools for real-time image alignment feedback are lacking.

Purpose of the Study:

• To develop and validate an automated algorithm for detecting medial/lateral probe translation errors in echocardiography.
• To compare the algorithm's performance against specialist assessments for image optimization.

Main Methods:

• A fully automated algorithm was developed using image analysis to detect incorrect ultrasound probe positioning.
• The algorithm's ability to differentiate optimal from sub-optimal four-chamber views was assessed.
• Algorithm performance was compared to consensus visual assessments by echocardiography specialists.

Main Results:

• The algorithm's orientation assessments showed strong correlation with specialist consensus ([Formula: see text]).
• Algorithm performance was comparable to the inter-specialist agreement ([Formula: see text]).
• The algorithm achieved 85% agreement with specialist consensus, with a mean discrepancy of [Formula: see text].

Conclusions:

• The automated algorithm reliably detects incorrect probe translation in echocardiography.
• This technology can provide real-time guidance for image optimization.
• It holds potential as a valuable tool for training and quality control in echocardiography.