Meeting Banner
Abstract #1754

Accuracy and Precision of DCE-MRI Parameters Estimated by AATH and MTK Models: Evaluations with MMID4 Simulation and Clinical NPC Datasets

Chen-Yi Liu1, Yen-Peng Liao2, Yu-Shi Lin1, 3, Shy-Chyi Chin4, Ho-Ling Liu1, 4

1Department of Medical Imaging and Radiological Sciences, Chang Gung University, Taoyuan, Taiwan; 2Department of Medical Imaging, Taipei Medical University - Shuang Ho Hospital, Taipei, Taiwan; 3Department of Diagnostic Radiology, Chang Gung Memorial Hospital, Keelung, Taiwan; 4Department of Medical Imaging and Intervention, Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan

This study aimed to investigate the variability of physiological parameters estimated by the AATH and the mTK models in DCE-MRI. Computer simulation were performed using multiple path, multiple tracer, indicator dilution, 4 region (MMID4) model with various noise added. In addition, the two models were applied in nasopharyngeal carcinoma patients for comparison. Computer simulation quantified the error and variations of the two models, which demonstrated that the AATH was more accurate but less precise than the mTK. Clinical results agreed with the simulation, which showed smaller Ktrans and Veand larger Vp with the AATH than with the mTK model.

Abstract PDF Poster PDF

Keywords

accuracy accurate acquisition acquisitions added adiabatic adjusting agreed aimed analyzed applied apply approximation arterial assess asset attempted back best better bioengineering blood breast capillary carcinoma characteristics chin clinical coefficient coil combined common commonly comparing computer concentration condition conditions consistent contrast converted crucial curve curves datasets dedicated degrees determined diagnostic dilution direct director dynamic enhanced equation especially essential evaluations experienced extend field fitting flow function gave general generating generation greater head homogeneity hospital increment indicator injection input interstitial interval intervention investigate iterations knowledge larger ling literature maps matrix medical memorial model modeling models modified national neck neutral neutrally noise noisy oncology parker particular path patient patients peak period permeability physiological plasma precise precision previous procedure product quantify radiological radiologist radiology recent repeated represented resource running sampling scanner sciences series simulated simulation sixty slice smaller status studies superior surface table tissue tracer true tumor uncertainties uncertainty uptake variance varied view volume widely years