Array Coils

2.1.1

Introduction

Essential parts of a magnetic resonance imaging (MRI) system, radiofrequency (RF) coils or antennas are used to transmit and/or receive signal. Phased-array coils combine multiple small coil elements to transmit or receive signal using independent channels. Array coils were first described in a seminal paper by Roemer et al., and proof-of-concept was demonstrated in the spine. Since then, array coils have become the standard for multiple body parts in research and clinical MRI. The two main advantages are their increased sensitivity and the possibility of faster acquisition via parallel imaging. Several concepts related to array coils need to be understood to appreciate their full benefits and optimize their use, notably the B ₁ magnetic field, the so-called geometric factor ( g -factor), and the signal-to-noise ratio (SNR).

This chapter will first introduce theoretical concepts related to RF coils, then will cover array coils and their advantages and limitations. We will review some of the existing array coils for human spinal cord imaging. We will then present design and building considerations, including safety aspects. Finally, we will look at how to evaluate transmit (Tx) and receive (Rx) coils (including the SNR and g -factor).

2.1.2

Coil Theory

The basic principle of RF coils is to create a B ₁ magnetic field that rotates the magnetization of the nuclei (Tx coils) or to measure signal emitted by the resonating nuclei in the transverse plane (Rx coils). Both Tx and Rx coils are tuned to the Larmor frequency ( $\omega$ ), at which particles can exchange energy. The Larmor frequency depends on the gyromagnetic ratio of the particle ( $\gamma$ ) and on the magnetic field strength they are submitted to ( B ), according to $\omega =\gamma B$ . For example at 3 tesla, the Larmor frequency of water protons is about 127 MHz.

Coils consist of a set of inductive ( L ) and capacitive ( C ) elements that are chosen to make the coil resonate at the Larmor frequency, according to:

Coils are designed for transmitting, receiving, or performing both functions. The geometry of the coil can be optimized for its purpose. Typically, for transmitting coils, B ₁ homogeneity should be high in order to obtain similar excitation (flip angle) over the region of interest. Conversely, for receiving, the ability to detect small magnetic flux generated by the resonating spins becomes more important than homogeneity, therefore sensitivity will be optimized. Thus, several coil geometries exist that achieve either good homogeneity or good sensitivity (sometimes both). Examples of volume and surface coils associated with their respective B ₁ field are shown in Figure 2.1.1 . Typically, volume coils are used for transmission since they are made up of large elements and can achieve good B ₁ homogeneity. Conversely, small surface coils are preferred for reception since they achieve high sensitivity. In standard MRI systems, the Tx coil is usually integrated into the scanner (also called the “body coil”), while the Rx coil is manually plugged into the system (either the patient's table or the body of the scanner).

To understand how receive coils work, let's consider a simple loop coil. According to Faraday's law, the alternating magnetic field produced by the nearby rotating nuclei induces an alternating current in the loop. This principle is similar to the bicycle dynamo from which a current is generated via a rotating small magnet nearby a solenoid coil. The sensitivity of a single loop can be obtained from the Biot–Savart law. The larger the loop, the deeper it can capture signal from, but also the higher the amount of noise it will capture from the object scanned. Hence, optimal loop size should be designed depending on the desired penetration depth. For example, to achieve optimal sensitivity at 8 cm deep, a loop of approximately 8 cm is desired.

2.1.3

Array Coils

2.1.3.1

Sensitivity

As mentioned in this chapter, a large loop can cover a large region but captures more thermal and physiological noise from the body. Contrariwise, a small loop captures less noise, as its sensitivity profile is restricted to a very small region close to the loop. The idea of having multiple small loops next to each other is to combine them in order to gain penetration in the sensitivity profile, while only capturing a small amount of noise coming from each individual coil element. This gives the rationale for designing coil arrays.

It is often mistakenly thought that in deep tissue, the sensitivity of phased-array coils with lots of small coils is poorer than that of coils with less but bigger elements. Although it is true that the penetration of each individual element of a phased-array coil is reduced, when combining several smaller elements into a phased-array coil, the sensitivity in deep tissue is similar to that of a coil with fewer elements. However, closer to the coil, the sensitivity of a phased-array coil is significantly higher. This was notably demonstrated using theoretical analyses and experiments. Figure 2.1.2 illustrates this point by comparing different coil arrangements.