Related Experiment Videos
Semi-automatic boundary detection to improve reporting of regional left ventricular function
J Timperley1, A R J Mitchell, D J Blackman
1Department of Cardiology, The John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. jon.timperley@orh.nhs.uk
Summary
A new semi-automatic boundary detection system improved reporting of regional left ventricular function for novice echocardiographers. This tool shows potential for enhancing training in cardiac imaging analysis.
Area of Science:
- Cardiology
- Medical Imaging
- Echocardiography
Background:
- Regional left ventricular function assessment relies on subjective interpretation of endocardial motion and thickening.
- This subjective method presents a significant learning curve for healthcare professionals.
Purpose of the Study:
- To evaluate if a semi-automatic boundary detection system can improve the reporting of regional left ventricular function.
- To assess the system's utility in standardizing assessments and aiding training.
Main Methods:
- A semi-automatic boundary detection system (Quamus) was used to generate images with superimposed endocardial borders.
- Echocardiographers of varying experience levels (Level 1, 2, and 3) assessed regional function using standard contrast images, images with Quamus borders, and Quamus borders alone.
Main Results:
- The addition of superimposed borders enhanced agreement among Level 1 echocardiographers, increasing weighted Kappa from 0.55 to 0.64.
- Agreement among Level 2 echocardiographers remained stable (0.63 to 0.64).
- The system demonstrated potential as a valuable training tool.
Conclusions:
- Semi-automatic boundary detection systems can aid in the objective assessment of regional left ventricular function.
- These tools may be particularly beneficial for less experienced practitioners and for training purposes in echocardiography.