Effect of automated image registration on radiologist interpretation

Abstract

In this study, we present preliminary data on the effect of automated 3D image alignment on the time to arrive at a decision about an imaging finding, the agreement of multiple of multiple observers, the prevalence of comparison examinations, and technical success rates for the image alignment algorithm. We found that automated image alignment reduced the average time to make a decision by 25% for cases where the structures are rigid, and when the scanning protocol is similar. For cases where these are not true, there is little or no benefit. In our practice, 54% of cases had prior examinations that could be automatically aligned. The overall benefit seen in our department for highly similar exams might be 20% for neuro and 10% for body; the benefit seen in other practices is likely to vary based on scanning practices and prevalence of prior examinations.

Fig 1

Arrangement of images across the four portrait-oriented PACS monitors. If there were more than seven series, the additional series would be displayed at the right side of the right-most monitor, and could be “dragged” to the desired display location. Ex 1 was always the “current” exam and Ex 2 was always the prior exam; the series numbers for the current examination were as scanned, and the prior examination was reordered to match the current.

Fig 2

Decision time for registered versus unregistered examinations, and for examinations with high similarity of series, and for low similarity of series. (a) Shows results for neuroradiology and (b) is for body images. Unreg = unregistered images, Reg = registered images. Similar examinations are those where the majority of series had the same contrast properties and scan plane as the current examination.

Fig 3

a Shows a single FLAIR image from the current and from a 6 months previous examination in a patient with an oligodendroglioma. Two of three observers indicated probable increase in size over the interval. b Shows the registered images as well as a subtraction image (subtraction was not provided for this study) which clearly demonstrates interval progression.