Image content extraction: Application to MR images of the brain

A system for automatically extracting image content features was developed that combines registration to a labeled atlas with natural language process ing of free-text radiology reports. The system was then tested with T1-weig hted, spoiled gradient-echo magnetic resonance (MR) imaging studies of the brain performed in nine patients. The locations of 599 structures were visu ally assessed by an experienced radiologist and compared with the locations indicated by automated output. The in-plane accuracy of the contours was s ubjectively evaluated as either good, moderate, or poor. The criterion for classifying a structure as correctly located was that 90% or more of all th e images containing the structure had to be correctly identified. For 98% o f the structures, the images identified by the automated algorithm agreed w ith those identified by the radiologist, and in 83% of cases, image contour s showed a good in-plane overlap. The results of this validation study demo nstrate that this combination of registration and natural language processi ng is accurate in identifying relevant images from brain MR imaging studies . However, the range of applicability of this technique has yet to be deter mined by applying the technique to a large number of studies.