Dual-Modality Imaging with SPECT/CT
Written in 2004
- Bruce Hasegawa, PhD
- Randall A. Hawkins, MD, PhD
Medical diagnoses commonly rely on assessment of both functional status and anatomical condition. Radionuclide techniques such as PET and SPECT provide physiologic and metabolic information but suffer from limited spatial resolution and lack of anatomic context. In contrast, CT and MRI provide excellent anatomic detail but limited functional insight.
Dual-modality imaging systems combine X-ray CT with PET or SPECT on a shared gantry, allowing co-registered acquisition of anatomical and functional data with the patient in the same position. This enables fused images where radionuclide distributions are overlaid on CT, improving disease localization, staging, and interpretation.
An additional benefit of dual-modality imaging is improved quantitative accuracy. CT-based attenuation maps allow correction of photon attenuation and scatter in radionuclide imaging, improving both localization and quantification. As a result, dual-modality imaging has gained widespread clinical adoption.
Clinical Implementation
- Dual-modality systems mount CT and radionuclide detectors on a single gantry with a shared patient table, allowing sequential CT and PET or SPECT scans with minimal repositioning.
- Image fusion is performed using registration software that accounts for differences in geometry and image matrix size between modalities.
- Early PET/CT and SPECT/CT prototypes were developed at academic centers, including UCSF and the University of Pittsburgh.
- Commercial systems were later introduced by GE, Siemens, CTI, and Philips/ADAC.
- Dual-modality systems for small-animal imaging have also been developed in academic and industrial settings.
- At UCSF, dual-modality imaging was pioneered by Christopher Cann, PhD; Robert Gould, ScD; Bruce Hasegawa, PhD; and Eric Gingold, PhD.
- UCSF research contributions included theoretical modeling, simulation, prototype scanner development, reconstruction methods, detector electronics, and ASIC development.
- These efforts enabled quantitative radionuclide uptake measurements in myocardial perfusion and cancer imaging models.
- The first human-scale clinical CT/SPECT system at UCSF combined a GE 600 XR/T SPECT camera with a GE 9800 Quick CT scanner.
- Subsequent studies evaluated myocardial perfusion, cancer imaging, and prostate cancer diagnosis using CT/SPECT.
- Ongoing work includes development of compact dual-modality systems for small-animal imaging.
Dual-Modality Imaging at UCSF
- GE Medical Systems and Elgems Ltd developed the Millenium VH “Hawkeye” SPECT/CT system, based on principles similar to those pioneered at UCSF.
- The Hawkeye system integrates SPECT, coincidence FDG imaging, and low-resolution CT for anatomical localization and attenuation correction.
- The system is installed at Moffitt Long Hospital and integrated with PACS and Nuclear Medicine workstations.
- It is used for routine clinical imaging including bone, renal, cardiac, and selected FDG coincidence studies.
- CT data acquired with the system improve attenuation correction and quantitative accuracy, especially in cardiac imaging.
- These capabilities support advanced research protocols, including radioimmunotherapy and in vivo dosimetry.
Optimizing PET-CT Quantification to Radiation Treatment Planning in Treating Cancers of the Head and Neck
Written in 2008
- Nick G. Costouros, MD
- Youngho Seo, PhD
- Stephen L. Bacharach, PhD
- Benjamin L. Franc, MD
- Randall A. Hawkins, MD, PhD
- Henry F. VanBrocklin, PhD
- Bruce H. Hasegawa, PhD
Cancer imaging traditionally relies on CT and MRI for anatomical definition but is limited in detecting early-stage disease and distinguishing tumor from post-treatment changes.
PET imaging with 18F-FDG enables detection of metabolically active tumors but can be confounded by inflammation and normal tissue uptake. To address these limitations, this project investigates 18F-fluorothymidine (FLT), a PET tracer associated with cellular proliferation, offering complementary biologic information.
Head and neck cancers are commonly treated with intensity-modulated radiotherapy (IMRT), which requires accurate definition of tumor volumes to maximize tumor dose while sparing critical structures.
Project Goals
- Develop novel PET image correction algorithms
- Derive biologic activity measures from PET-CT data
- Improve display formats for incorporating biologic tumor volumes into IMRT planning
The goal is to validate PET-based biologic imaging for radiation planning compared with conventional anatomy-based approaches.
Methods
- Patients with head and neck cancer underwent sequential FDG and FLT PET-CT imaging prior to surgery.
- Imaging was performed using standardized immobilization and a Siemens Biograph 16 PET-CT scanner.
- Dynamic imaging and whole-body PET-CT scans were acquired and reconstructed using consistent algorithms.
- Standardized uptake values and Patlak kinetic analyses were calculated using region-of-interest and arterial input functions.
- PET-derived metrics were correlated with pathology findings including tumor presence, Ki-67 index, and nuclear-to-cytoplasmic ratios.
Results
- FDG and FLT PET-CT imaging was successfully performed in three patients.
- Increased uptake was observed in primary tumors, metastatic lymph nodes, and reactive lymph nodes.
- Pathology confirmed elevated proliferation indices corresponding to PET findings.
- Overlap between tumor and reactive lymph node uptake highlighted the need for pathology-based thresholds.
Conclusion
- Quantitative variability in PET-CT limits its current use in radiation treatment planning.
- Integrating biologically meaningful PET metrics with CT anatomy could significantly improve IMRT accuracy.
- Voxel-level biologic mapping overlaid on anatomical CT holds promise for reducing morbidity and improving outcomes.
- Further data are required to establish robust biologic thresholds for clinical implementation.
Acknowledgment: NIBIB T32 Training Grant T32 EB001631; UC Discovery Grant l04-10174 with Siemens Medical Solutions; NCI Grant K25 CA114254.