Olga A. Kubassova PhD,MSc, Presenter: Founder, Image Analysis Ltd Director, Image Analysis Ltd

Mikael Boesen PhD, Abstract Co-Author: Advisor, Image Analysis Ltd Speaker, Esaote SpA

Mikkel Ostergaard, Abstract Co-Author: Nothing to Disclose

Henning Bliddal MD, PhD, Abstract Co-Author: Nothing to Disclose

Marco Amedeo Cimmino MD, Abstract Co-Author: Nothing to Disclose

Mette Bjorndal Axelsen MD, Abstract Co-Author: Nothing to Disclose

Rene Panduro Poggenborg MD, Abstract Co-Author: Nothing to Disclose

Rasmus Bouert, Abstract Co-Author: Nothing to Disclose

Anshul Rastogi MBBS, FRCR, Abstract Co-Author: Consultant, Image Analysis Ltd

Nikolay Tzaribachev, Abstract Co-Author: Nothing to Disclose

Mark Hinton, Abstract Co-Author: Employee, Image Analysis Ltd

Peter C. Taylor BMBCh, FRCR, MA, PhD, Abstract Co-Author: Advisor, Image Analysis, Inc

Synovitis is an early indicator of inflammatory disease activity. It is visualised in Dynamic Contrast Enhanced MRI (DCE-MRI) and quantified by assessing the height and slope of time vs. intensity curves. This study investigates the robustness of DCE-MRI Quantification method (DEMRIQ) for assessment of early inflammatory changes in rheumatoid arthritis (RA) patients.

Three independent studies were performed to quantify DCE-MRI of wrist and knee: 1) 3T DCE-MRIs of wrist acquired from 10 healthy patients 4 times over a year; 2) 1T DCE-MRIs of wrist acquired in 26 healthy controls and 14 early RA patients under treatment over a year; 3) 12 1.5T DCE-MRI of RA knee joints twice over 6 months. The protocols were standardised to use GRE sequences acquired every 10-12 seconds over 5-6 min. The Initial Rate of Enhancement (IRE), Maximum Enhancement (ME) and the total number of enhancing voxels were automatically calculated with Dynamika (Image Analysis, UK) from rough ROIs drawn by two independent readers around the anatomy. DEMRIQ-Volume, the sum of voxels with plateau and wash-out inside the ROI and DEMRIQ-Inflammation, the mean of IRE inside the ROI multiplied by DEMRIQ-Volume, were calculated and compared with the state of the art scoring for RA MRI - RAMRIS. Spearman’s rank correlation (ρ) was calculated between the results of both scoring methods. Inter Class Correlation (ICC) coefficients were calculated between the scores of the observers.

Correlation between DEMRIQ and RAMRIS was ρ>0.865, p<0.05. Longitudinal changes of ME and IRE were stable in controls, under 0.04 compared to baseline values. Healthy controls values were lower than baseline RA values for all parameters (Mann-Whitney, p<0.005). IRE decreased during treatment (Wilcoxon signed rank test, p<0.005), showing sensitivity to change. ICC=0.95, p<0.005.

DEMRIQ is robust to different scan parameters, correlates well with RAMRIS and allows for continuous assessment and high degree of automation and reproducibility. The method has the potential to become a sensitive marker for detecting early and subtle changes.

Automation of DCE-MRI quantification allows for objective and reproducible decision support in clinical research and diagnosis. DEMRIQ allows for continuous assessment as opposed to discrete scores of 0-3 with RAMRIS, leading to much more personalised approach to treatment management and earlier diagnosis.

Kubassova, O, Boesen, M, Ostergaard, M, Bliddal, H, Cimmino, M, Axelsen, M, Poggenborg, R, Bouert, R, Rastogi, A, Tzaribachev, N, Hinton, M, Taylor, P, DEMRIQ: A Dynamic Contrast Enhanced MRI Quantification Method for Objective Assessment of Treatment Response in Patients with Inflammatory Arthritis.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14006643.html