A geometric index to differentiate abnormal from normal septal wall motion on cardiac MRI

Research output: Contribution to journalArticlepeer-review


Background The definition of abnormal septal wall motion (SWM) is usually performed on a subjective visual assessment with cardiac MR (CMR). Purpose To quantify SWM using a geometric index and to compare this index to the visual inspection in patients with or without abnormal SWM. Material and Methods Cine CMR images of 100 consecutive patients were retrospectively assessed for visual evaluation of SWM and calculation of a convexity septal index (CSI) on mid-ventricular short-axis images, defined as b/a (a = minimal distance from anterior to posterior ventricular junctions; b = maximal distance from septum to a), obtained in end-systole (ES) and end-diastole (ED). Cohen κ, Bland-Altman method, Kruskall-Wallis, Mann-Whitney U, Jonckheere-Terpstra, and Spearman statistics were used to compare the two methods. Results At visual evaluation, 73 patients had normal SWM (group A), nine abnormal SWM at ES (group B), and 18 at ES and ED (group C). Median CSI for group A was significantly higher (ES = 0.23, ED = 0.25) than those for group B and C (0.10-0.15) (P <0.001). Inter-reader reproducibility of visual evaluation was high (κ = 0.841, P <0.001) while that for CSI was 77%. Average time for visual analysis was 3 min, for quantitative analysis 4 min. Conclusion CSI is a simple and reproducible way to quantify SWM. ED CSI seems to be sensitive in detecting abnormal SWM in patients with apparently normal SWM at visual evaluation.

Original languageEnglish
Pages (from-to)545-551
Number of pages7
JournalActa Radiologica
Issue number5
Publication statusPublished - May 1 2015


  • Cardiac MR (CMR)
  • convexity septal index (CSI)
  • septal wall motion (SWM)

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology
  • Medicine(all)


Dive into the research topics of 'A geometric index to differentiate abnormal from normal septal wall motion on cardiac MRI'. Together they form a unique fingerprint.

Cite this