57 A feasability study for radiographer reporting and analysis of left ventricular ejection fraction in cardiac magnetic resonance imaging

Abstract

Introduction The purpose of this study was to compare the accuracy of a reporting radiographer to an experienced cardiologist in reporting MRI assessment of Left Ventricular Ejection Fraction (LVEF) on patients taking potentially cardiotoxic trastuzumab based chemotherapy (including Herceptin). Radiographer reporting of MRI studies has become more and more widespread (Snaith et Al 2015) However, to date there have been limited instances of radiographer reporting cardiac MRI studies. All Herceptin studies are currently reported by consultant cardiologists and radiologists. It was proposed that an experienced MRI radiographer could be trained to report these studies as a way of relieving pressure upon the cardiac service. Method 100 sequential Herceptin MRI studies were performed on a Philips Ingenia 1.5 tesla MRI scanner. The lead cardiac radiographer with 8 years of cardiac MRI experience was trained to use Medis suite MR version 3.2 (published by Medis medical imaging systems) software package by a senior cardiologist. The appointment time for each patient was 30 minutes with 15 minutes of actual scanning time. The sequences used were: Black blood axial stack. Balanced Gradient echo Cine sequences performed in 2 chamber(CH), 4CH and 3CH planes with a short axis cine stack (SAX) prescribed through the entirety of left ventricle. LV volume and EF were measured from the SAX. Each study was reported separately by an imaging consultant and by the radiographer who were blind to each-others reports. Results Results are shown in figure 1 below Of the 100 studies reported: 95 had a measured EF difference of 2% or lower between reporters. Of the remaining 5, only 1 had an EF difference greater than 4%. (see figure 1) The higher percentage difference in this case was 13%, due to the fact that the patient had a prosthetic mitral valve in situ, artefact from which rendered measurements unreliable. This case was evaluated by three other imaging consultants (blinded to the original results) and the EF measurements ranged from 44% to 62%. Conclusion We have shown that a trained and experienced MRI radiographer calculates LVEF with low levels of variability. (+/- 2% in 95% of cases). This compares favourably with published data comparing the reproducibility of LVEF measurements between imaging consultants (Bhuva et al 2019, George 2017) The difficulty in assessment of LV function in the presence of a prosthetic Mitral valve is widely accepted (Gulsin et al 2017, Shellock 2016) and resulted in an extreme outlier in the results. The consequent variation in opinion between imaging consultants is evidence that this discrepancy was due to the presence of significant artefact form the prosthesis. The reporting radiographer is now awaiting trust approval to commence independently reporting Herceptin cases. This study has also highlighted the potential for radiographer reporting in other areas of cardiac MRI which we intend to explore. Conflict of Interest none