Dosimetry in 131I internal emitter therapy using voxel dependent integrated time-activities derived from multiple, registered SPECT and CT images

The accuracy of dosimetry calculations in internal emitter therapy applications is often limited by a lack of data describing the time dependence of the spatial activity concentration in patients. In many cases, time-dependent activity data is acquired only for specific regions of interest (ROIs), and measurements of patient anatomy and 3D activity distributions are made at only single time points during tracer studies. When calculating absorbed dose in these applications, this scarcity of data necessitates the approximation that activity distributions can be wholly defined by single spatial measurements, and that all points within given ROIs exhibit uniform time dependence. In the current work, CT and SPECT images have been acquired using a dual-modality scanner at multiple time points after administration of both tracer and therapy activity to follicular lymphoma patients being treated with ¹³¹I tositumomab. The data has been registered to a single CT image, and the mutually registered SPECT images have been used to derive integrated time-activities on a voxel-by-voxel basis. Maps of integral time-activity have been used in conjunction with CT images to determine 3D absorbed dose distributions by Monte Carlo computation. Results are presented illustrating the differences between calculations of spatial distributions of integrated activities and absorbed doses made with the current technique and those performed with previous methods.