Simulated Multi-Protocol CT Imaging with Patient-Derived Numerical Phantom Towards Patient-Specific Scan Protocol Optimization
Z Yang*, H Jin, J Kim, Seoul National University, Seoul, KoreaSU-E-I-20 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Optimizing scan protocols for individual patients and exam of interest is clinically demanding, and yet challenging task.
This study presents a simulated CT imaging technique with patient-derived numerical phantom which enables exploration of combination of protocols that might lead to a condition for dose-effective and acceptable image quality without repeated scanning patient with multiple protocols.
From a wholebody patient CT scan, 24 different tissue types were segmented semi-automatically. Spectral attenuation properties of tissue types were referenced from NIST, and natural state density of each tissue was referenced from the literature. X-ray spectra of various KVp were generated by using an existing TASMIP model. Projection/filtered-backprojection process in fan-beam geometry taking quantum noise into account were computed to generate simulated CT images for any combination of KVp, mAs, and slice thickness. In order to mimic real CT images, filter kernels were derived from measured MTFs of a commercial CT(Sensation 16, Siemens). Developed technique was used for generating multi-protocol CT images of 5 different KVp (120, 110, 100, 90, 80 KVp) and varying mAs to assess the feasibility of patient-specific optimization of CT protocol that might lead to minimum achievable dose while preserving diagnostic image quality. For quality evaluation, CNR was calculated on ROIs placed on liver and adipose tissues.
Noise and texture appearance of the simulated CT image resembled sufficiently to the real CT images from commercial CT system. Multiple combinations of KVp and mAs were found to yield comparable CNR(10.7 ~ 11.5) while reducing patient dose by 25%.
Proposed method could generate simulated multi-protocol CT images of patient-derived phantoms, which has a potential for use in optimizing the patient-specific scan protocol.
Funding Support, Disclosures, and Conflict of Interest: The research was supported by the Converging Research Center Program through the Ministry of Education, Science and Technology (2012K001498)