MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel
  1. Home
  2. Meeting Program
  3. All Sessions
  4. All Imaging General ePoster Discussion
  5. Evaluating Image Quality in MR K-Space Substitution Technique

Evaluating Image Quality in MR K-Space Substitution Technique

S Deng2*, A Zhao1 , J Vasquez1 , G Clarke1,2 , (1) Dept of Radiology, University of Texas Health San Antonio, San Antonio, TX, (2) Research Imaging Institute, University of Texas Health San Antonio, San Antonio, TX,

Presentations

(Wednesday, 7/17/2019) 10:30 AM - 11:00 AM

Room: Exhibit Hall | Forum 8

Purpose: Accuracy of T1w signal and change in resolution in k-space substitution (e.g. Keyhole imaging) can be evaluated using ADNI phantom. Aims: (i) compare T1w phantom signal error fluctuation as k-space substitution increases (ii) investigate how k-space change the noise level and (iii) investigate to what extent spatial resolution is affected.

Methods: An ADNI phantom (EMR051, Magphan®, containing eight T1 inserts) was placed in the head coil of 3T Siemens TIM Trio MRI system. Two sets (n=4) of T1W-MPRAGE images were collected using different inversion times (TI=1500,930,1500,1100 ms respectively) with TR=3.6s, bandwidth=180kHz, no acceleration and flip angle=6. Keyhole technique was applied in phase encoding (PE) direction and frequency (FE) encoding direction to generate different percentage of k-space substitution. FFT and sum of square were applied to generate coil-combined magnitude images. Precision error of the eight T1W phantoms signals, and noise profiles were evaluated with varying percentage k-space substitution. Spatial limiting resolution was evaluated using modulation transfer function. Consistency of measurement errors was evaluated using the standard error of mean (SEM).

Results: The signal errors started to be consistently low for 60 lines replaced in PE direction (0.8%,SEM 0.1% and lower thereafter), and for 100 lines in FE direction (1.3%, SEM 0.4%, and lower thereafter, Figure 1). Substitution in FE direction gives higher T1W signal error than in PE direction (5.3±14.2% vs 1.7±4.2% group mean/SD), while with lower noise(159.6±33 vs 197.1±89). Limiting resolution varies from 0.294–0.282 for PE (3.4% in variance), and 0.268-0.254 for FE (8% in variance) with different substitutions.

Conclusion: T1 error is consistently low with 60/256 or more of k-space replaced in PE direction, while more varying errors are found in FE directions. Fluctuations in spatial resolution and noise also were identified, but at an acceptable range.

Keywords

Not Applicable / None Entered.

Taxonomy

IM- MRI : Quantitative imaging/analysis

Contact Email

Additional Links