EVALUATION OF THE LIMITS OF VISUAL DETECTION OF IMAGE MISREGISTRATIONIN A BRAIN F-18 FLUORODEOXYGLUCOSE PET-MRI STUDY

In routine clinical work, registration accuracy is assessed by visual inspection. However, the accuracy of visual assessment of registration has not been evaluated. This study establishes the limits of visual d etection of misregistration in a registered brain fluorine-18 fluorode oxyglucose positron emission tomography to magnetic resonance image vo lume, The ''best'' registered image volume was obtained by automatic r egistration using mutual information optimization. Translational movem ents by 1 mm, 2 mm, 3 mm and 4 mm, and rotational movements by 1 degre es, 2 degrees, 3 degrees and 4 degrees in the positive and negative di rections in the x- (lateral), y- (anterior-posterior) and z- (axial) a xes were introduced to this standard, These 48 images plus six ''best' ' registered images were presented in random sequence to five observer s for visual categorization of registration accuracy. No observer dete cted a definite misregistration in the ''best'' registered image. Eval uation for inter-observer variation using observer pairings showed a h igh percentage of agreement in assigned categories for both translatio nal and rotational misregistrations. Assessment of the limits of detec tion of misregistration showed that a 2-mm translational misregistrati on was detectable by all observers in the x- and y-axes and 3-mm trans lational misregistration in the z-axis. With rotational misregistratio ns, rotation around the z-axis was detectable by all at 2 degrees rota tion whereas rotation around the y-axis was detected at 3-4 degrees. R otation around the x-axis was not symmetric with a positive rotation b eing identified at 2 degrees whereas negative rotation was detected by all only at 4 degrees. Therefore, visual analysis appears to be a sen sitive and practical means to assess image misregistration accuracy. T he awareness of the limits of visual detection of misregistration will lead to increase care when evaluating registration quality in both re search and clinical settings.