February 2017
S M T W T F S
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28
IIIT MK Newsletter Subscription
Name:

Email:

PDF Print

DEVELOPMENT AND VALIDATION OF SIGNAL PROCESSING TECHNIQUES FOR TIME-CONSTRAINED MAGNETIC RESONANCE IMAGING (DST SR/S3/EECE/0107/2012)

The objective is to enhance the contrast in MR image when the acquisition time is constrained to be limited in comparison to those of available MR sequences. Harmonic retrieval methods will be used for synthesizing the k-space from sparsely sampled and incompletely acquired spin/gradient echo signal. A significant outcome is the development of a partial-echo sequence, in which the acquisition time per RF pulse application can be limited to 1.2-1.5TE for both spin and gradient echo sequences. A linear prediction-based algorithm embodies the principle underlying the possibility of recovering anatomical details of the proposed partial-echo sequence. Prior to clinical trials, the prototypical sequence and the associated image reconstruction method will be validated using an MR simulator, and anatomically realistic simulated data sets. The remainder of outcomes is related to clinical applications involving brain lesion identification and fat-water separation using the proposed partial-echo sequence.

SIGNAL PROCESSING FOR RAPID AND PARTIAL ECHO-PLANAR ACQUISITION IN MAGNETIC RESONANCE IMAGING OF NEURAL ISCHEMIC STROKE (ETP/18/2013/KSCSTE).

Diffusion-weighted MRI (DWI) and Fluid Attenuated Inversion Recovery (FLAIR) form a combinational approach in the investigation of stroke imaging. DWI derived measures are used to detect the infarcted brain tissue and FLAIR images are used to detect regions with edema. Of these, the DWI employs Echo-Planar Imaging (EPI) type of acquisition in which the entire k-space is scanned with a single RF pulse, followed by echoes whose amplitudes are modulated in accordance with the distance from the k-space center. EPI data have inherently low Signal-to-Noise Ratio (SNR). In situations where only partial echoes are acquired, the low SNR of EPI data can further result in loss of fine structural details. In the case of rapid scanning, echo truncation in EPI will need extra processing to restore the lost information. Our method to investigate the incomplete (partial) acquisition of EPI and bring out corrective filtering steps to improve quantitative stroke imaging. The key research components include development of spatial and k-space filters for elimination of truncation artefacts, enhancement and restoration of features in the magnitude image reconstructed from partial k-space of EPI data. As compared to other multi-pulse sequences, the filters required for restoration of partially acquired EPI are to be tailored for operation under low SNR conditions. A novel aspect of the current proposal is the elimination of the need for acquiring a major fraction of the echoes in the dephasing period.

Last Updated on Monday, 01 September 2014 10:59