Program Information
Quantification of Short TE Breast MRSI with Combined SVD and AMERES Method
W Feng1, A Chu2 , J Hu3* , (1) Bayhealth Medical Center, Tenafly, NJ, (2) St. Jude Children's Research Hospital, Memphis, TN, (3) Wayne State University, Detroit, MI
Presentations
MO-L-GePD-I-6 (Monday, July 31, 2017) 1:15 PM - 1:45 PM Room: Imaging ePoster Lounge
Purpose: The lipid in breast is a mixture with various lipid chain lengths, which showed as different peaks in MRS; unfortunately total number of lipid type and concentration are unknown, it is not possible to generate all individual spectrums, so no standard lipid basis are available. Simulated lipid spectrums can be generated as basis set, however QUEST or AQSES cannot reliably quantify the Cro from short TE breast data. We combine the SVD method to filter the large water, lipid and macromolecules peaks, and then use AMERES to quantify the Cro peak from the residue FID.To study the Quantification method of Breast short TE MRS by combine AMERES and SVD Method.
Methods: Patient breast MRSI acquired with Siemens Verio 3T scanner with TR 1500ms and TE 80ms, pulse sequence includes optional inversion recovery, optional outer volume presaturation (OVP), and an echo-filter MRSI with a weighted k-space sampling scheme. Hankel SVD method is used to compute signal poles and amplitude, from them signal frequencies and damping factors. Based on frequency and damping factor, large water, lipid and macromolecules peaks are filtered first, residue FID is used for quantification with AMERES.
Results: Fileted peaks include 4.6-4.7 ppm water peaks, 1.3, 1.2, 0.9, 1.7, 1.1, 5.2, 3.8, 2.4, 2.16 lipid peaks, which correspond to known lipid peaks 1.3, 0.9, 1.6, 2.02, 2.23, 2.75, 4.1, 4.3, 5.2, 5.3 ppm within 0.1ppm, except 2.4 and 3.8ppm. AMERES can quantify Chro peak around 3.2 ppm, however AMERES cannot handle negative amplitude.
Conclusion: As demonstrated, combined SVD filter method and AMERES method can be used to quantify short TE breast MRS data, results are satisfactory for acquired 3T patient data.
Contact Email: