Medical College of Wisconsin
CTSICores SearchResearch InformaticsREDCap

Phase-aligned multiple spin-echo averaging: a simple way to improve signal-to-noise ratio of in vivo mouse spinal cord diffusion tensor image. Magn Reson Imaging 2014 Dec;32(10):1335-43



Pubmed ID


Pubmed Central ID




Scopus ID

2-s2.0-84961318352   8 Citations


PURPOSE: To improve signal-noise-ratio of in vivo mouse spinal cord diffusion tensor imaging using-phase aligned multiple spin-echo technique.

MATERIAL AND METHODS: In vivo mouse spinal cord diffusion tensor imaging maps generated by multiple spin-echo and conventional spin-echo diffusion weighting were examined to demonstrate the efficacy of multiple spin-echo diffusion sequence to improve image quality and throughput. Effects of signal averaging using complex, magnitude and phased images from multiple spin-echo diffusion weighting were also assessed. Bayesian probability theory was used to generate phased images by moving the coherent signals to the real channel to eliminate the effect of phase variation between echoes while preserving the Gaussian noise distribution. Signal averaging of phased multiple spin-echo images potentially solves both the phase incoherence problem and the bias of the elevated Rician noise distribution in magnitude image. The proposed signal averaging with Bayesian phase-aligned multiple spin-echo images approach was compared to the conventional spin-echo data acquired with doubling the scan time. The diffusion tensor imaging parameters were compared in the mouse contusion spinal cord injury. Significance level (p-value) and effect size (Cohen's d) were reported between the control and contused spinal cord to inspect the sensitivity of each approach in detecting white matter pathology.

RESULTS: Compared to the spin-echo image, the signal-noise-ratio increased to 1.84-fold using the phased image averaging and to 1.30-fold using magnitude image averaging in the spinal cord white matter. Multiple spin-echo phased image averaging showed improved image quality of the mouse spinal cord among the tested methods. Diffusion tensor imaging metrics obtained from multiple spin-echo phased images using three echoes and two averages closely agreed with those derived by spin-echo magnitude data with four averages (two times more in acquisition time). The phased image averaging correctly reflected pathological features in contusion spinal cord injury.

CONCLUSION: Our in vivo imaging results indicate that averaging the phased multiple spin-echo images yields an 84% signal-noise-ratio increase over the spin-echo images and a 41% gain over the magnitude averaged multiple spin-echo images with equal acquisition time. Current results from the animal model of spinal cord injury suggest that the phased multiple spin-echo images could be used to improve signal-noise-ratio.

Author List

Tu TW, Budde MD, Xie M, Chen YJ, Wang Q, Quirk JD, Song SK


Matthew Budde PhD Associate Professor in the Neurosurgery department at Medical College of Wisconsin

MESH terms used to index this publication - Major topics in bold

Bayes Theorem
Diffusion Tensor Imaging
Echo-Planar Imaging
Image Processing, Computer-Assisted
Mice, Inbred C57BL
Normal Distribution
Signal-To-Noise Ratio
Spinal Cord
Spinal Cord Injuries