Combined Dual-Layer Imager and Fast Kv-Switching Improves Linac Spectral Onboard Imaging
Abstract
Purpose
Spectral imaging confers several benefits such as material decomposition, increased soft tissue contrast, and metal artifact reduction. Novel methods currently under evaluation for adoption to linac onboard imaging include a kV dual-layer imager investigated by our group and fast kV-switching of the linac kV source. This study investigates potential benefits from the combination of these two methods to yield four different energy channels.
Methods
A script was written to toggle a linac’s kV source from 125kVp to 80kVp during a conebeam CT (CBCT). A dual-layer imager (DLI) was mounted to the linac and each layer was read out separately. Comparisons were made with a CBCT, 125kVp DLI only, 125-80kVp kV-switching only (top layer only), and max separation (bottom layer DLI 125kVp, top layer DLI 80kVp). Virtual monoenergetic images (VMI) were created. A CatPhan was scanned to assess linear attenuation coefficient accuracy for the different inserts with an 80keV VMI. Contrast-to-noise ratio (CNR) was measured on a thorax phantom with implanted lung tumor and a pelvis phantom’s prostate with 50keV VMIs. Metal inserts replaced the pelvis phantom’s femoral heads for an additional scan to assess metal artifact reduction using 150keV VMIs.
Results
All CatPhan linear attenuation coefficients were within 2% of nominal values. Max separation increased CNR by 20.7% over CBCT and 9.9% over standard spectral imaging (thorax tumor with lung background) and by 96.3% over CBCT and 41.5% over standard (prostate with adipose background). With metal inserts adjacent muscle uniformity for max separation increased 112.9% over CBCT and 36.1% over standard.
Conclusion
Fast kV-switching was successfully combined with a DLI on a linac’s onboard imaging system. The combined approach outperforms standard kV imaging as well as kV-switching or DLI alone. Future work will combine all four energy channels in VMI generation.