Stretchable coil arrays: application to knee imaging under varying flexion angles.

To fit high-density receiver arrays for MRI closely around individual target anatomies, there is a need to provide a high degree of geometric adjustability with ease of handling and patient comfort. In this work, this is accomplished by the construction of a coil array that is stretchable such that it automatically conforms to a given anatomy's shape and size. Stretchability is implemented by creating the coil conductors from braided wire mounted on an elastic textile substrate. The signal-to-noise ratio yield of such coils is measured by MRI experiments at 3 T, and the signal-to-noise ratio effect of coil stretching is investigated with and without adjustment of the matching between each coil and the respective preamplifier. Four-channel and eight-channel arrays of stretchable receiver coils are evaluated in phantoms as well as for in vivo imaging of the human knee. Exploiting stretchability, it is demonstrated that the knee can be imaged under varying flexion angles up to 60° while maintaining closely coupled array detection, high signal-to-noise ratio, and uniform coverage of the entire joint.

Noise figure characterization of preamplifiers at NMR frequencies.

A method for characterizing the noise figure of preamplifiers at NMR frequencies is presented. The noise figure of preamplifiers as used for NMR and MRI detection varies with source impedance and with the operating frequency. Therefore, to characterize a preamplifier's noise behavior, it is necessary to perform noise measurements at the targeted frequency while varying the source impedance with high accuracy. At high radiofrequencies, such impedance variation is typically achieved with transmission-line tuners, which however are not available for the relatively low range of typical NMR frequencies. To solve this issue, this work describes an alternative approach that relies on lumped-element circuits for impedance manipulation. It is shown that, using a fixed-impedance noise source and suitable ENR correction, this approach permits noise figure characterization for NMR and MRI purposes. The method is demonstrated for two preamplifiers, a generic BF998 MOSFET module and an MRI-dedicated, integrated preamplifier, which were both studied at 128MHz, i.e., at the Larmor frequency of protons at 3 Tesla. Variations in noise figure of 0.01dB or less over repeated measurements reflect high precision even for small noise figures in the order of 0.4dB. For validation, large sets of measured noise figure values are shown to be consistent with the general noise-parameter model of linear two-ports. Finally, the measured noise characteristics of the superior preamplifier are illustrated by SNR measurements in MRI data.

Mechanically adjustable coil array for wrist MRI.

In this work, the concept of mechanically adjustable MR receiver coil arrays is proposed and implemented for the specific case of human wrist imaging. An eight-channel wrist array for proton MRI at 3 Tesla was constructed and evaluated. The array adjusts to the individual anatomy by a mechanism that fits a configuration of flexible coil elements closely around the wrist. With such adjustability, it is challenging to ensure robust electrical behavior and signal-to-noise (SNR) performance. These requirements are met by preamplifier decoupling and a suitable matching strategy based on pi networks that render the coil responses robust against changes in tuning, loading and mutual coupling. The robustness of the resulting SNR yield was studied by varying the effective coil matching over a wide range in a phantom imaging experiment. While SNR variation of up to 25% was observed at the surface of the phantom the SNR was essentially constant in the critical center region. A second SNR study in wrist phantoms of different sizes confirmed the benefits of bringing the coil elements very close, up to 3 mm, to the individual target volume. These findings were supported by initial in vivo imaging, exploiting high-sensitivity detection for highly resolved structural imaging.