Dinesh Thavendiranathan and Vamshi Kotha want to invite you to participate....

Looking for CMR Readers of all abilities to:

  • Evaluate LV Volume, EF and Mass according to current guidelines on a supplied, standardized data set
  • Complete a survey on your education and training 

The data generated will help create a training tool to standardize measurements across Canada

More detailed information follows below...... 

Project Name:  

Assessment of Improvement in Inter-Observer Variability and Accuracy of LV contouring on CMR Images through a group-based intervention using current guidelines and recommendations

Project Description:

Despite CMR being considered a gold standard for LV analysis, there is currently very little data on the variability amongst multiple readers in a CMR laboratory and even less so between readers from different CMR laboratories. Although recommendations have been published on suggested methods for post-processing there can still be significant inter-observer variability in interpreting individual patient images. Studies have shown that there is some improvement in agreement between readers with different levels of training after the implementation of consensus guidelines but these have mostly been single centre efforts. Further, the lack of an external reference standard meant that the analyses could not ensure that improvements in inter-observer variability did not come at the cost of accuracy.

The objectives of our proposed study are to (1) assess the inter-observer variability and accuracy of LV volume, EF, and mass assessment using CMR amongst multiple readers from several CMR laboratories across Canada, (2) assess intra-observer variability among a select group of senior CMR readers, (3) determine if a group review and educational intervention can improve the inter-observer variability and maintain or improve the accuracy amongst readers, (4) build a composite dataset of LV measurements to facilitate training of new readers and to be used in the development of a pan-Canadian benchmark, (5) correlate variation from the composite dataset of each reader with the length of training and practice by way of a survey and (6) identify a subgroup of individuals that may benefit the most from such an intervention. The latter is particularly important as performing calibration exercises amongst all members of a laboratory can be challenging.  Therefore targeting these interventions to those who would benefit the most would make such an effort more feasible.

The strengths of this undertaking would be the inclusion of several CMR laboratories across Canada and several CMR readers with different levels of training at each laboratory, use of a robust case base across the range of LV functions and the presence of a same-day MUGA as an external reference standard to assess deviation from baseline mean, ensuring accuracy.




Multi-Centre, Inter-Observer Study:  LV Quantification of Mass and Function