General Design Features of Dual Modality Imaging Systems

Modern dual-modality imaging systems incorporate subsystems for radio-nuclide imaging and for x-ray computed tomography that essentially use the same components as those in dedicated nuclear medicine and CT systems. In PET/CT, the radionuclide detector uses a scintillator (bismuth germinate, lutetium orthosilicate, gadolinium orthosilicate) coupled to an array of photomultiplier tubes for imaging the annihilation photons from the positron-emitting radiopharmaceuticals. Modern PET/CT scanners also include an x-ray source and detector identical to those used in modern multislice helical scanning CT scanners.33-35 Similarly, SPECT/CT systems use conventional dual-headed scintillation cameras suitable for planar scintigraphy or tomographic imaging of single-photon radionuclides, or coincidence-imaging of PET radiopharmaceuticals. The first-generation clinical SPECT/CT scanners used a low-resolution CT detector40,41,70,105 that offered relatively modest scan times (i.e., approximately 20 seconds per slice). However, newer SPECT/CT scanners now are becoming available that incorporate state-of-the-art multislice helical CT scanners identical to those used for diagnostic CT procedures.

The integration of the radionuclide and x-ray imaging chains in a dual-modality imaging system requires special considerations beyond those needed for scanners designed for single modality imaging alone. One challenge is offered by the presence of x-ray scatter from the patient that has the potential to reach and possibly damage the radionuclide detectors which are designed for the relatively low photon fluence rate encountered in radio-nuclide imaging.34'35'64 To avoid this possibility, the radionuclide detector in a dual-modality system typically is offset in the axial direction from the plane of the x-ray source and detector. This distance can be relatively small when a modest x-ray tube is used such as the 140 kV, 1 mA tube used in the GE Millennium VG SPECT/CT system,64 but can be 60 cm or more when a diagnostic CT scanner is coupled to a modern PET scanner operated without septa.28,30,31,34

As noted above, all dual-modality systems rely on separate x-ray and radionuclide imaging chains that must be supported on a common mechanical gantry to maintain consistent spatial relationship between the two data sets, and allow the detectors to be rotated and positioned accurately for tomographic imaging. The requirements for translational and angular positioning accuracy are, of course, different for CT, SPECT, and PET. For example, CT requires approximately 1000 angular samples acquired with an angular position and center of rotation maintained with submillimeter accuracy. In comparison, SPECT and PET have spatial resolutions of a few millimetres, and therefore can be performed with an accuracy of slightly less than a millimetre for clinical imaging.

The mechanical gantry of the dual-modality imaging system obviously must be designed to satisfy the requirements for both the radionuclide image and for CT. This can be achieved in several ways. In first generation SPECT/ CT systems,40,41,70,105 the SPECT detectors and CT imaging chain were mounted on the same rotating platform and were used sequentially while rotated around the patient. This limited the rotational speed of the x-ray and radionuclide imaging chains to approximately 20 sec per rotation, but also had the advantage that it could be performed using a gantry similar to that used with a conventional scintillation camera. Second generation SPECT/ CT systems that now are available include high-performance diagnostic CT subsystems. This requires the heavy SPECT detectors to be mounted on a separate rotating platform from the CT imaging chain which is rotated at speeds of 0.25 to 0.4 sec per revolution. While this design obviously increases the performance of the CT subsystem, it also increases the cost and complexity of the gantry.

In comparison to SPECT/CT in which the radionuclide detector is rotated around the patient during data acquisition, PET typically (but not always) is performed using a dedicated high-end stationary detector ring. A PET/CT system therefore can be configured by designing a gantry that mounts a stationary PET detector ring in tandem with a platform that rotates the CT imaging chain around the patient using a mechanical configuration similar to that used in a conventional diagnostic CT scanner. Alternatively, a partial ring of PET detectors can be rotated to acquire the PET data using the same rotating platform as the CT subsystem. This approach was taken by Town-send and his colleagues in their implementation of the first PET/CT sys-tem,28,30,31 and is an alternative for a more economical dual-modality system in comparison to those that use a full-ring of PET detectors. All of these mechanical designs have been used in commercial dual-modality systems and obviously offer trade-offs in terms of their performance and cost.

The patient table is another seemingly simple, yet important element of a dual-modality scanner.34,35 Most imaging systems use cantilevered patient tables to support the patient in the bore of the imaging system. Patient tables are designed to support patients weighing up to 500 pounds, but obviously deflect to varying degrees when they are loaded and extended with normal adult patients. However, dual-modality systems use x-ray and radionuclide imaging chains in tandem and thereby require longer patient tables than conventional imaging systems. Moreover, the table extension and the degree of deflection can be different for the x-ray and radionuclide imaging chains which can introduce a patient-dependent inaccuracy in the registration of the x-ray and radionuclide images. This problem is overcome by several different methods. The first uses a patient table which is supported in front of the scanner, with a secondary support between or at the far end of the x-ray and radionuclide imaging chains to minimize table deflection. A second approach adopted by CTI Molecular Imaging and Siemens Medical Systems uses a patient table that can be fixed on a base that is translated across the floor to extend the patient into the scanner. Since the patient platform is stationary relative to its support structure (which acts as a fulcrum), the deflection of the patient table is identical when the patient is positioned in the radionuclide imager or the CT scanner.

Finally, in modern dual-modality scanners, the computer systems are well integrated in terms of system control, data acquisition, image reconstruction, image display, and data processing and analysis.34,106 The dual-modality system must calibrate the CT data so that it can be used as an attenuation

map to correct the radionuclide data for photon attenuation. ^ ^ For physician review, the dual-modality system also typically registers the CT and radionuclide data and presents the radionuclide image as a colour overlay on the grey-scale CT image. Finally, software tools are provided that, for example, allow a cursor placed on the CT image by the operator with another cursor automatically placed in the identical position on the radionuclide image, and vice versa. These software functions allow the operator to utilize the dual-modality data in correcting, viewing, and interpreting and obviously are important design elements in modern dual-modality imaging systems.

0 0

Post a comment