Summary

MR Physics: Summary

Signal processing tools are fundamental to MRI, from the design of acquisition strategies through to the analysis of the resulting images. Linear systems theory and digital filter design are used to spatially localize the signal of interest through selective excitation. Data acquisition occurs in the spatial frequency (k-space) domain where sampling theory determines resolution and field of view. Strategies for reducing image artifacts are often best developed in this domain. Image reconstruction involves fast multi-dimensional Fourier transforms, often preceded by data interpolation, re-sampling, and apodization. The resulting large digital data sets lend themselves to computer image analysis. Considerations for image analysis either performed by the computer or an expert observer begin with appropriate selection of acquisition parameters; there is a great deal of flexibility in trading off SNR and resolution and maximizing contrast between tissues of interest at this stage. As the quality of the data continues to improve with better hardware, imaging sequences, and analysis tools, new clinical applications are being developed for this very flexible, non-invasive modality. With these applications come greater challenges for the underlying signal processing.