The Program in Non-Fourier and RF-Encoding began with the implementation of wavelet-encoded MRI by Panych in 1993  followed by the introduction and implementation of SVD-encoded MRI by Panych and his long-time collaborator, Gary Zientara . Wavelet and SVD encoding are among the earliest examples of an approach known as “non-Fourier encoded MRI” [1-5,7-14]. Unlike direct imaging methods such as X-ray in which spatial information is acquired directly (in the image domain) and thus require no special reconstruction, data in MRI is "encoded". The basic type of encoding in MRI is Fourier encoding, a direct consequence of using imaging gradients that create a linear relationship between spatial location and frequency of the MRI signal.
Non-Fourier encoding takes advantage of the ability to selectively excite regions of the imaging volume with tailored radio-frequency (RF) pulses. For this reason, it is also sometimes referred to as “RF encoding”. With the flexibility of selectively exciting specific regions, it is possible to encode MR data with any arbitrary set of spatial functions or patterns . The earliest work in non-Fourier encoding perfected the technology of encoding via RF selective excitation and advancing the theory supporting the method.
From the initial work in non-Fourier encoding, a new dynamic imaging methodology known as “dynamically adaptive imaging” made its appearance [3,8,15]. The goal in dynamically adaptive imaging is to utilize the information obtained by processing image data on-line - - ideally, in real time - - in order to optimize the data acquisition. In theory, non-Fourier encoding should be more suited to adaptive imaging than Fourier encoding because of the flexibility to employ arbitrary encoding schemes that allow for encoding functions to be tailored to the application. For example, wavelet encoding was explored for implementing multi-resolution adaptive approaches that involved zooming into regions of change . SVD encoding was found to be particularly interesting as it aimed to adapt encoding functions dynamically as the contents of the imaging volume changed [8,19-20].
Efforts in the development of the theory and technology of non-Fourier encoding were soon complemented by work on applications such as functional MRI, interventional MRI and others. Projects on the use of non-Fourier encoding and adaptive imaging for functional MRI were led by Seung-Schik Yoo, a doctoral student, and a postdoctoral fellow [11,15,18]. Lei Zhao, another student, investigated the application of non-Fourier encoding for hyperpolarized gas imaging [16,38]. Non-Fourier encoding, and particularly adaptive imaging, were first proposed with interventional MRI in mind and over the years a number of projects have focused on this application [17,28,32]. Due to practical considerations, however, approaches have tended to be directed more toward hybrid methods combining RF encoding with standard Fourier encoding rather than the pure non-Fourier encoding as it was initially envisioned. Methods such as MURPS [17,28] and PSF-Choice [22,29] were developed and applied along with hybrid reduced field-of-view imaging [6, 21, 23-27].