Routine clinical use of image coregistration remains limited in spite of numerous applications for enhanced interpretation and analysis of medical images. The clearest benefit exists for improved interpretation of serial images to assess disease progression. This includes a critical need for objectively monitoring the effect of established and potential interventions or therapy on diseases such as tumors and multiple sclerosis. Further, putting together various types of image data with intermodality (or multimodality) coregistration provides information not available from analysis of the individual modalities. Initial work from the medical centers investigating use of coregistration in clinical settings has already demonstrated the benefit of multimodality coregistration in assessing and directing treatment of brain tumor recurrence and localizing epilepsy for surgical treatment. Presurgical epilepsy localization also includes the use of coregistration to combine electrophysiologic epilepsy localization with functional imaging of brain mapping to determine the safest and most effective surgical strategies with the fewest of invasive procedures. Applying intersubject coregistration is probably the farthest away from routine clinical use, but it offers a true advance for accurate quantitative interpretation of functional anatomic imaging.
Reasons for slow adoption of image coregistration in the clinical setting are increasingly difficult to support. Many of the reasons for not taking advantage of coregistration have been mainly a result of logistical difficulty in efficiently implementing automated computerized manipulation of heterogeneous image data. Deployment of high-speed networks in most hospitals is removing the obstacle of quick access to digital image data archives. Validated coregistration techniques that are accurate enough for clinical purposes are also now widely available for most applications. All that remains is for clinicians and imaging departments to work more closely at identifying specific needs for clinical image coregistration, and then to simply collect, adapt, and employ existing image processing tools.
Was this article helpful?