Automated vs Manual Cardiac MRI Planning: Reliability and Scan Time Comparison

Overview

This prospective study compared AI-based automated cardiac MRI planning with manual planning in terms of error rates and scan duration. Automated planning demonstrated potential to reduce planning errors and streamline scan times while maintaining protocol consistency.

Background

Cardiac magnetic resonance imaging (CMR) is a complex diagnostic tool requiring extensive technologist training due to the need for precise cardiac plane identification and sequence parameter adjustments. Manual planning is time-consuming and prone to errors, which can prolong scan times and reduce image quality. Automated planning software aims to fully automate CMR acquisition steps, potentially reducing human error and improving efficiency. Prior studies have explored partial automation, but comprehensive clinical evaluation of fully automated CMR planning remains limited.

Data Highlights

The study enrolled consecutive patients undergoing non-stress CMR, randomized weekly between automated and manual planning arms. Automated planning utilized a vendor-specific AI software integrating auto-positioning, auto-alignment, and automatic parameter adjustments. Manual planning included additional phase-contrast sequences not present in the automated protocol. Examinations were performed on a single 1.5-T scanner with technologists of varying experience levels. Patient compliance, heart rate, rhythm, and breathing strategy were recorded to assess their impact on scan performance.

Key Findings

• Automated CMR planning software successfully performed all steps required for protocol execution with minimal human intervention.
• Automated planning reduced the likelihood of planning errors compared to manual approaches, potentially improving dataset homogeneity.
• Scan times were decreased in the automated group, reflecting improved procedural efficiency.
• Technologist cognitive load was reduced by automating repetitive parameter settings, allowing greater focus on image quality assessment.
• Automated planning maintained protocol consistency despite patient variability in heart rate, rhythm, and breathing strategy.

Clinical Implications

Implementing fully automated CMR planning can enhance workflow efficiency by reducing scan duration and minimizing operator-dependent errors. This approach may improve patient comfort and image quality consistency, facilitating longitudinal clinical assessments and standardized research protocols. Clinicians and technologists should consider integrating AI-driven planning tools to optimize CMR examinations.

Conclusion

Automated cardiac MRI planning demonstrates promise in reducing errors and scan times while maintaining protocol fidelity, representing a significant advancement in CMR workflow optimization. Further studies may validate its broader clinical applicability.

References

1. Siemens Healthineers AG -- myExam Cardiac Assist and AI Cardiac Scan Companion
2. Prior Research [14,15] -- AI-based Automated CMR Planning Framework