Low-field magnetic resonance imaging (LFMRI) is a form of medical imaging that uses a system of magnets with field strengths of approximately 0.25 to 1 Tesla (T) to generate detailed images of tissue structures. Traditional clinical MRI systems usually operate at a range of 1.5 to 3 T or higher, since the signal-to-noise ratio (SNR), a key metric of image quality, is roughly proportional to magnetic field strength. However, advances in technology have enabled the production of clinically useful images at significantly lower field strengths.
Background Unlike high-field MRI systems that require cryogenic cooling, extensive radio frequency (RF) shielding, and heavy infrastructure, low-field MRI scanners can be built using permanent or resistive magnets, reducing cost, complexity, and power requirements. These systems are often smaller, lighter, and more portable, making them suitable for point-of-care imaging, emergency or bedside use, and deployment in resource-limited or remote environments. LFMRI therefore has many global health applications, especially in low- and middle-income countries (LMICs), where conventional MRI is often unavailable due to its expense and logistical demands. Traditional MRI scanners were developed in the late 1970s, and these first models used resistive electromagnets below 3 T. However, in the 1980s, as the technology developed, the resistive magnets were replaced with superconductor magnets due to the homogeneity of the magnetic fields and the high field strengths, which allowed for higher image quality. However, since 2018, several LFMRIs have been approved by the FDA for use in clinical settings. These scanners have been used in emergency departments, for neonatal imaging, and for portable diagnostic settings. Overall, LFMRI has the potential to expand the accessibility of MRI images and high-quality diagnostic healthcare across the globe.
Technical principles of traditional MRI
MRI is based on the behavior of hydrogen nuclei in strong external magnetic fields. The applied magnetic field aligns a portion of the hydrogen protons in the body, creating a net magnetization that can be measured to create the image.
Gradients and slice selection MRI relies on two key magnetic field components: the strong static field B0 and the gradient fields that vary linearly in space. The B0 field, produced by the main magnet, aligns hydrogen protons and sets a uniform Larmor frequency throughout the scanner. To localize signals and form images, MRI systems apply gradient fields along the x, y, and z axes, slightly altering the magnetic field strength in different regions of the body. Because the Larmor frequency depends on magnetic field strength, these gradients cause proton precession frequencies to vary with position. Slice selection uses this principle by applying an RF pulse simultaneously with a gradient field (along the z axis or along the body head to toe), so that only spins whose Larmor frequency falls within the narrow bandwidth of the RF pulse are excited. By choosing the gradient orientation and RF bandwidth, the system can excite a single slice of tissue at a specific z-axis location, enabling spatially resolved imaging in three dimensions. The Larmor frequency can be described by the equation ω0 = γB0, where γ is the gyromagnetic ratio specific to hydrogen protons, and B0. This means that the Larmor Frequency increases linearly with magnetic field strength. High-field MRI systems therefore, operate at comparatively high radio frequency (RF) ranges, which enhances the strength of the detectable signal.
Image generation The RF pulse tips the protons in the specific slice into alignment with the pulse (away from the B field). Once the RF pulse is turned off, the protons begin to "relax" back toward equilibrium. The recovery of magnetization along the direction of B0 is called longitudinal relaxation, characterized by the T1 relaxation time. The changing magnetic flux induces voltages in receiver coils, allowing the MRI system to detect the returning signal. In addition to T1 recovery, protons also undergo transverse dephasing, described by the T2 relaxation time. After excitation, the spins precess in unison in the transverse plane, generating a strong detectable signal. However, interactions between neighboring spins cause them to spin out of sync over time, leading to a decay of the transverse magnetization. This process is represented by the T2 relaxation constant, which is similarly picked up by receiver coils.
Image reconstruction Once relaxation signals are measured by the receiver coils, the MRI system must convert them into an image. The signal data are first encoded in k-space, a frequency–spatial domain that represents how the signal varies with the applied gradient fields. Each z-slice determined by RF slice selection has its own k-space data, allowing for spatially localized information for the image. After sufficient data are acquired, mathematical operations, primarily the Fourier transform, are used to convert the k-space data into a spatial image. The resulting image reflects how T1, T2, and proton density vary across tissues, enabling detailed visualization of anatomical and physiological structures.
Equipment specifications of traditional MRI Traditional MRI systems rely on superconducting magnets, which provide highly uniform and stable magnetic fields. These magnets are cooled with liquid helium to maintain superconductivity, enabling systems to operate with minimal resistive losses. However, the need for cryogenic cooling, RF shielding, and extensive site preparation makes these scanners large, expensive, and stationary. Despite the logistical demands, superconducting magnets remain the standard in clinical imaging due to the high field strengths and image quality they enable.
LFMRI system design and components
Magnet system At the center of an LFMRI scanner is a magnet that establishes a static magnetic field B0. LFMRI systems use either permanent magnets or resistive electromagnets. High-strength NdFeB permanent magnets are most commonly used because of their strength and stability.
… excerpt ends here. Continue reading the full article.
