Using Human and Model Performance to Compare MRI Reconstructions
Authors:
- M Dylan TisdallMTM Dylan Tisdall
- M.S. AtkinsMAM.S. Atkins
- 30 October 2006
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 25 (11) , 1510-1517
Abstract
Magnetic resonance imaging (MRI) reconstruction techniques are often validated with signal-to-noise ratio (SNR), contrast-to-noise ratio, and mean-to-standard-deviation ratio measured on example images. We present human and model observers as a novel approach to evaluating reconstructions for low-SNR magnetic resonance (MR) images. We measured human and channelized Hotelling observers in a two-alternative forced-choice signal-known-exactly detection task on synthetic MR images. We compared three reconstructions: magnitude, wavelet-based denoising, and phase-corrected real. Human observers performed approximately equally using all three reconstructions. The model observer showed very close agreement with the humans over the range of images. These results contradict previous predictions in the literature based on SNR. Thus, we propose that human observer studies are important for validating MRI reconstructions. The model's performance indicates that it may provide an alternative to human studiesKeywords
This publication has 16 references indexed in Scilit:
- Efficiency of the human observer detecting random signals in random backgroundsJournal of the Optical Society of America A, 2005
- Noise reduction for magnetic resonance images via adaptive multiscale products thresholdingIEEE Transactions on Medical Imaging, 2003
- A wavelet-based method for improving signal-to-noise ratio and contrast in MR imagesMagnetic Resonance Imaging, 2000
- Effect of image compression in model and human performance1999
- Objective assessment of image quality III ROC metrics, ideal observers, and likelihood-generating functionsJournal of the Optical Society of America A, 1998
- Noise in MRIMagnetic Resonance in Medicine, 1996
- The rician distribution of noisy mri dataMagnetic Resonance in Medicine, 1995
- Effect of random background inhomogeneity on observer detection performanceJournal of the Optical Society of America A, 1992
- Homodyne detection in magnetic resonance imagingIEEE Transactions on Medical Imaging, 1991
- Addition of a channel mechanism to the ideal-observer modelJournal of the Optical Society of America A, 1987