(TeslaFlow) Evaluation of Automatic Deep-Learning Based Slice Prescription for Cardiac MRI

Conventional cardiac MRI acquisitions involve time-consuming and operator-dependent prescription of imaging planes including a short-axis (SAX) stack and multiple long-axis (LAX) planes. Therefore, extensive anatomic and technical expertise is necessary, and is a major reason why cardiac MRI is predominantly offered only at tertiary care centers with special expertise.
TeslaFlow allows automatic MR slice prescription using deep learning algorithms. It holds promise to provide accurate, automated, and reproducible prescription of MR acquisitions, irrespective of patient position and technologist experience. In doing so, it promises to optimize workflow and efficiency.
TeslaFlow advertises reproducible planning to ensure exam consistency for same patient follow-up irrespective of patient position. However, there have been no prospective studies embedded in an active clinical environment to determine the “real world” performance of the TeslaFlow approach. This study plans to explore this.