The Bias of Hardware and Software Studies in Complex Image Summation around a Spherical or Cylindrical Object (Cissco) Method

CISSCO is a quantitative method to obtain magnetic moment and susceptibility of objects in MRI. While magnetic moment and susceptibility quantifications of small objects in-vivo using CISSCO method has been extensively investigated, the bias of CISSCO method and image artifacts from the factors of multichannel receiving coil, the processing of multichannel coils, parallel imaging, and background field removal algorithms across MRI platforms, are relatively unexplored. Clinical setups have often appeared these factors. This two-year research proposal will study the impacts of each factor on signal-to-noise (SNR), image artifacts, and the bias of CISSCO method. The optimal or compromised options to reduce the bias of CISSCO and image artifacts for these factors will be provided in this proposal as well. At the first year, a gel phantom with cylindrical and spherical objects will be built and scanned using multi-echo gradient echo sequence in MRI scanners with and without these factors. The images and results of quantified magnetic moment and susceptibility in CISSCO will reveal image artifacts and the bias, and the optimal or compromised options will be provided. At the second year, the study will focus on fine tuning between SNR and resolution in numerical simulations to achieve reasonable SNR and relatively high resolution using the optimal or compromised options. With this fine tuning relation and suggested options, normal human subjects will be scanned in MRI scanners. These discrepancies between results in phantom and human studies with and without using these factors will be reported and minimized. These results in this proposal will lead to better image quality, less image artifacts and the reproducible results in CISSCO method. These results will also move a step forward standardization of susceptibility quantifications. In addition, initiating conversation and discussion with MRI vendors based on these findings, the parameters of hardware and software would be adjusted to reduce image artifacts. Beside, these findings have implications for the research and clinical groups which utilize phase information to diagnose or tackle diseases.

Project ID:PB10703-0461
External Project ID:MOST106-2218-E182-004