Table of Contents

1. What is an MRI Scanner?
2. Why is an MRI Scanner Used?
3. How Does an MRI Scanner Work?
4. What Are the Main Components of an MRI Scanner?
5. What Types and Variants of MRI Scanners Exist?
6. What Are the Main Benefits of an MRI Scanner?
7. What Are the General Risks and Limitations of MRI Scanners?
8. How Is MRI Technology Evolving? Recent Innovations
9. Key Takeaways and Tips for Biomedical Engineers

1. What is an MRI Scanner?

Definition and Core Concept

Magnetic Resonance Imaging (MRI) is a non-ionizing, cross-sectional medical imaging modality that exploits the quantum mechanical property of nuclear spin — specifically of hydrogen protons abundant in biological tissue — to generate high-resolution anatomical and functional images. Unlike X-ray machines or CT scanners, MRI produces no ionizing radiation, instead relying on powerful magnetic fields and radiofrequency (RF) energy to interrogate tissue properties. The result is exquisite soft-tissue contrast that renders MRI indispensable across virtually every clinical specialty.

A Brief History of MRI Development

The scientific foundations of MRI trace back to Felix Bloch and Edward Purcell, who independently demonstrated nuclear magnetic resonance (NMR) in 1946 and shared the Nobel Prize in Physics in 1952. The clinical leap came in 1971 when Raymond Damadian showed that tumour tissue exhibited markedly different NMR relaxation times compared to healthy tissue. In 1973, Paul Lauterbur introduced the concept of using magnetic field gradients to encode spatial information, while Peter Mansfield contributed the mathematical framework for rapid image acquisition — work that earned them the Nobel Prize in Physiology or Medicine in 2003. By the early 1980s, commercial whole-body scanners were entering hospitals, and the technology has since evolved from low-field resistive magnets to today’s superconducting systems operating at 1.5 T, 3 T, and beyond.

Basic Physical Principle

At its core, MRI exploits the Larmor precession of hydrogen proton spins within a strong, static magnetic field (B₀). When placed in B₀, proton spins align either parallel or anti-parallel to the field, establishing a net longitudinal magnetization. A precisely tuned RF pulse at the Larmor frequency (f₀ = γ·B₀, where γ = 42.58 MHz/T for ¹H) tips this magnetization into the transverse plane, inducing a measurable signal as the spins subsequently relax back to equilibrium. Spatial localization is achieved through superimposed magnetic field gradients, enabling reconstruction of two- and three-dimensional images with millimetre-scale resolution.

2. Why is an MRI Scanner Used?

Clinical Indications

MRI has become the gold standard for imaging the central nervous system, musculoskeletal structures, abdomen, pelvis, and cardiovascular system. Key clinical indications include detection and characterisation of brain tumours, stroke assessment (diffusion-weighted imaging), multiple sclerosis lesion mapping, spinal cord pathology, cartilage and ligament injury, liver fibrosis staging, prostate cancer localisation, and cardiac function evaluation. Functional MRI (fMRI) maps neural activation patterns, while MR angiography (MRA) evaluates vascular anatomy without iodinated contrast. The breadth of diagnostic information obtainable from a single examination, without radiation exposure, makes MRI an exceptional tool in both cancer diagnosis and treatment planning and routine clinical care.

Advantages Over Other Imaging Modalities

Compared to CT, MRI offers superior soft-tissue contrast without ionizing radiation, making it preferable for repeated imaging and paediatric patients. Relative to ultrasound imaging systems, MRI is not limited by acoustic windows, bone shadows, or operator dependency, and provides whole-organ volumetric data. Nuclear medicine techniques such as PET offer metabolic information but lack the spatial resolution and anatomical detail achievable with MRI. The multiparametric nature of MRI — generating T1-weighted, T2-weighted, proton density, diffusion, perfusion, and spectroscopic contrast within one examination — is unmatched by any other single modality.

Role in Modern Diagnosis and Treatment Guidance

Beyond pure diagnosis, MRI serves as a real-time guidance platform for interventional procedures such as MR-guided focused ultrasound (MRgFUS), biopsy, and neurosurgery. Intraoperative MRI (iMRI) systems allow surgeons to visualise residual tumour during resection. In radiation oncology, MRI-guided radiotherapy (MR-Linac) enables adaptive treatment planning that accounts for daily anatomical variation. As AI-driven reconstruction and analysis tools mature, MRI’s role in quantitative biomarker extraction and personalised medicine continues to expand rapidly.

3. How Does an MRI Scanner Work?

Step 1 — Static Magnetic Field (B₀) and Spin Alignment

When a patient is placed inside the MRI bore, the strong static field B₀ causes hydrogen proton magnetic moments to align along the field axis, splitting into parallel (low energy) and anti-parallel (high energy) populations. The slight excess in the parallel state produces a small but detectable net magnetisation vector (M₀) proportional to B₀ and proton density. Higher field strengths yield a larger M₀ and therefore inherently greater signal-to-noise ratio (SNR), which is a primary driver behind the clinical migration toward 3 T systems and the research push to 7 T and beyond, as explored in detail in our article on how MRI technology and AI are reshaping diagnostics.

Step 2 — Radiofrequency Excitation and the Rotating Frame

A short burst of RF energy delivered at the Larmor frequency resonantly tips M₀ away from the longitudinal axis by a flip angle α (typically 90° for spin-echo sequences). In the rotating reference frame, the magnetisation precesses in the transverse plane, inducing a time-varying electromotive force in a receiver coil — this constitutes the free induction decay (FID) signal. The amplitude, frequency, and phase of the FID encode tissue properties and spatial location. Critically, the energy deposited in tissue must be controlled to prevent heating; the Specific Absorption Rate (SAR), measured in W/kg, is monitored in real time and must comply with IEC 60601-2-33 limits (e.g., ≤2 W/kg averaged over the whole body for normal operating mode).

Step 3 — T1 and T2 Relaxation Contrast Mechanisms

Following excitation, magnetisation recovers through two independent relaxation processes. Longitudinal (T1) relaxation describes regrowth of M₀ along B₀ as protons exchange energy with the surrounding molecular lattice; fat has a short T1 (~250 ms at 1.5 T) and appears bright on T1-weighted images, while cerebrospinal fluid has a long T1 (~4,000 ms) and appears dark. Transverse (T2) relaxation reflects dephasing of the transverse magnetisation due to spin-spin interactions; fluids exhibit long T2 values and appear bright on T2-weighted images, facilitating oedema detection. Additionally, T2* relaxation incorporates local magnetic field inhomogeneities and underpins blood-oxygen-level-dependent (BOLD) contrast used in fMRI. Radiographers manipulate repetition time (TR) and echo time (TE) to preferentially weight images toward T1, T2, or proton density contrast.

Step 4 — Gradient Spatial Encoding and Image Reconstruction

Three sets of gradient coils superimpose linearly varying magnetic fields onto B₀ in the x, y, and z directions. Slice-select gradients confine excitation to a defined anatomical plane. Frequency-encoding gradients impose a spatial frequency shift along one in-plane axis during signal readout, while phase-encoding gradients step through successive phase shifts along the perpendicular in-plane axis between TR intervals. The acquired data, stored as a two-dimensional matrix of spatial frequencies called k-space, is transformed via the Fast Fourier Transform (FFT) to yield the final image. Advanced acquisition strategies such as echo-planar imaging (EPI), parallel imaging (SENSE, GRAPPA), and compressed sensing dramatically accelerate k-space filling, reducing scan times from minutes to seconds.

4. What Are the Main Components of an MRI Scanner?

Superconducting Magnet and Cryogenic System

The superconducting magnet is the dominant subsystem by cost, weight, and complexity. Niobium-titanium (NbTi) alloy wire wound into a solenoid coil is cooled to approximately 4 K using liquid helium, reaching a superconducting state in which electrical resistance vanishes and persistent current flows indefinitely without an external power source. Modern zero-boil-off systems use closed-cycle cryocoolers (Gifford-McMahon or pulse-tube refrigerators) to recondense helium vapour, dramatically reducing helium consumption. The magnet assembly incorporates passive and active shielding coils to confine the fringe field, minimising the 5-Gauss exclusion zone required for patient and equipment safety. Field homogeneity across the imaging volume must be better than a few parts per million (ppm), achieved through an active shimming system comprising room-temperature resistive shim coils and superconducting shim elements.

Radiofrequency Coils

RF coils serve dual roles as transmitters (generating the B₁ excitation field) and receivers (detecting the NMR signal). The body coil, integrated into the bore housing, provides uniform whole-body transmission. Dedicated surface and phased-array receive coils — head coils, spine coils, cardiac coils, knee coils, and others — are placed close to the anatomy of interest to maximise SNR. Phased-array coils with 32, 64, or even 128 independent channels enable parallel imaging acceleration by exploiting the distinct spatial sensitivity profiles of each element. High-field systems require careful RF coil design to manage B₁ field inhomogeneity and dielectric effects, often employing multi-channel transmit (pTx) architectures for RF shimming.

Gradient Coils and Power Electronics

Gradient coils consist of precisely wound resistive conductors that generate linear field variations in x, y, and z. Performance is characterised by maximum gradient amplitude (mT/m, typically 40–80 mT/m on clinical systems) and slew rate (T/m/s, up to 200 T/m/s), which govern spatial resolution and sequence speed, respectively. Gradient amplifiers supply switching currents of thousands of amperes with microsecond precision. Acoustic noise — a well-known patient discomfort — arises from Lorentz forces acting on current-carrying gradient windings in the strong B₀ field; acoustic dampening enclosures and quiet-gradient pulse design techniques mitigate this. Rapidly switching gradients can also induce peripheral nerve stimulation (PNS), a physiological safety limit governed by IEC 60601-2-33.

Computer, Reconstruction, and Operator Console Systems

The MRI system computer orchestrates precise synchronisation of RF pulse generation, gradient switching, and data acquisition. Raw k-space data are transferred to high-performance reconstruction engines — increasingly GPU-accelerated — that execute FFT-based or iterative reconstruction algorithms in near real time. The operator console provides a graphical interface for protocol selection, patient registration, scan geometry planning, and image review. Modern scanners integrate AI-assisted reconstruction pipelines such as Siemens Deep Resolve, GE AIR Recon DL, and Philips SmartSpeed, which apply deep-learning denoising to enable dose-efficient, high-SNR imaging at reduced scan times. DICOM-compliant image management and PACS integration complete the data workflow, aligning with quality management requirements under ISO 13485.

5. What Types and Variants of MRI Scanners Exist?

Overview of Scanner Configurations

MRI scanners are categorised primarily by magnet configuration and field strength, each presenting distinct trade-offs among SNR, patient accessibility, imaging speed, cost, and siting requirements. Major commercial platforms include the Siemens MAGNETOM series, GE SIGNA, Philips Ingenia, Canon Vantage, and Hitachi Echelon, spanning field strengths from 0.5 T to 7 T. Understanding these variants is essential for biomedical engineers involved in procurement, facility planning, equipment management, or device development — competencies highlighted among the top skills every biomedical engineer should master.

Emerging Portable and Point-of-Care MRI

A transformative development is the emergence of low-field portable MRI systems (e.g., Hyperfine Swoop, 0.064 T), which use permanent or resistive magnets and require no dedicated RF-shielded room. These devices enable bedside neuroimaging in ICUs, emergency departments, and resource-limited settings. While SNR and resolution remain inferior to high-field systems, AI-based image enhancement is rapidly narrowing the clinical utility gap. Regulatory classification of such devices involves careful navigation of FDA frameworks, as discussed in our overview of how biomedical devices are classified.

Comparison of MRI Scanner Types

The table below summarises the principal MRI scanner variants, providing a concise reference for biomedical engineers evaluating system specifications for clinical or research applications:

6. What Are the Main Benefits of an MRI Scanner?

Diagnostic Advantages

MRI scanners offer unparalleled soft tissue contrast, making them indispensable for diagnosing neurological disorders, musculoskeletal injuries, abdominal pathologies, and oncologic conditions. Unlike CT or X-ray imaging, MRI differentiates between tissues of similar density by exploiting differences in proton density and relaxation times (T1, T2, and T2*). Multiplanar imaging capability — axial, coronal, sagittal, and oblique planes — is achievable without repositioning the patient, providing comprehensive anatomical coverage in a single examination. For biomedical engineers interested in the broader landscape of diagnostic imaging modalities, the article on Radiological Devices: All You Must Know as a Biomedical Engineer provides essential complementary context.

Patient Safety Benefits

One of the most clinically significant advantages of MRI is the complete absence of ionizing radiation. Unlike X-ray, fluoroscopy, or CT, MRI relies solely on radiofrequency energy and magnetic fields, eliminating cumulative radiation dose concerns. This makes MRI the modality of choice for pediatric patients, pregnant women (particularly in the second and third trimesters), and individuals requiring repeated longitudinal imaging — such as patients undergoing cancer surveillance. The technology is also inherently well-suited to advancing cancer diagnosis and treatment, where frequent monitoring is essential without the risks associated with accumulated radiation exposure.

Versatility and Functional Imaging

Beyond structural anatomy, MRI enables a rich spectrum of functional and physiological assessments. Functional MRI (fMRI) measures blood-oxygen-level-dependent (BOLD) signals to map brain activation patterns, supporting both neuroscience research and pre-surgical planning. Cardiac MRI delivers high-resolution assessment of myocardial function, viability, and perfusion — a key tool for biomedical engineers specializing in cardiovascular devices. MR spectroscopy (MRS) provides biochemical profiling of tissues by detecting metabolites such as choline, creatine, and N-acetylaspartate. Diffusion-weighted imaging (DWI) and perfusion imaging extend diagnostic capability further, enabling early stroke detection and tumor characterization.

7. What Are the General Risks and Limitations of MRI Scanners?

Safety Hazards and Contraindications

The strong static magnetic field of an MRI scanner poses significant risks to patients and staff who have ferromagnetic implants or foreign bodies. Pacemakers, cochlear implants, certain aneurysm clips, and metallic orbital fragments may experience torque, translational force, or device malfunction in the magnetic environment. Projectile accidents involving ferromagnetic objects (the so-called “missile effect”) represent serious safety hazards in the scanner room. Specific absorption rate (SAR) limits govern radiofrequency energy deposition to prevent tissue heating, with IEC 60601-2-33 defining whole-body SAR thresholds of 2 W/kg for normal operating mode and 4 W/kg for first-level controlled mode. Gadolinium-based contrast agents, while generally safe, carry risks of nephrogenic systemic fibrosis (NSF) in patients with severely compromised renal function, and are subject to increasing scrutiny regarding long-term tissue deposition. Understanding these safety constraints is critical for engineers involved in regulatory submissions, as detailed in our guide on How Biomedical Devices Are Classified: Insights into FDA Classification.

Technical and Operational Limitations

MRI acquisition times remain considerably longer than CT or X-ray, typically ranging from 20 to 60 minutes for a full examination. Long scan times increase susceptibility to motion artifacts and reduce patient throughput. Acoustic noise generated by rapidly switching gradient coils can reach levels as high as 110 dB, necessitating hearing protection for patients and posing design challenges for engineers developing quieter gradient systems. Metal artifact reduction sequences (MARS) and advanced reconstruction algorithms partially mitigate image degradation near orthopedic implants, but significant challenges persist. The high capital cost — approximately $1–3 million for a 1.5T system and $2–3 million for a 3T system — along with substantial infrastructure requirements (shielding, cryogen supply, HVAC) represent formidable barriers to accessibility, particularly in low- and middle-income settings.

Patient-Related Challenges

Claustrophobia affects an estimated 5–10% of patients, sometimes preventing examination completion even with anxiolytics. Wide-bore and open-configuration MRI systems have partially addressed this, though open MRI typically operates at lower field strengths with reduced image quality. Pediatric and critically ill patients frequently require sedation or general anesthesia, introducing additional procedural risk and staffing requirements. Patients with severe obesity may exceed table weight limits (typically 200–250 kg) or bore diameter constraints. These human factors present ongoing engineering and clinical design challenges that biomedical engineers must consider during system selection and facility planning.

8. How Is MRI Technology Evolving? Recent Innovations

AI and Deep Learning Reconstruction

Artificial intelligence is fundamentally transforming MRI acquisition and reconstruction pipelines. Compressed sensing (CS) accelerates data acquisition by undersampling k-space and recovering diagnostic-quality images through iterative reconstruction algorithms that exploit sparsity. Deep learning-based reconstruction networks — such as Siemens Deep Resolve, GE AIR Recon DL, and Philips SmartSpeed — train convolutional neural networks on large datasets to denoise and reconstruct images from significantly reduced raw data, enabling scan time reductions of up to 50–70% without measurable loss of diagnostic quality. AI-powered automated segmentation, lesion detection, and workflow prioritization tools are further reducing radiologist workload. For biomedical engineers seeking to understand this transformative intersection, the article on Latest Advances of Artificial Intelligence in Healthcare in 2025 offers an authoritative overview.

Ultra-High Field Systems and Silent MRI

The clinical approval of 7T MRI systems — including the Siemens MAGNETOM Terra and Philips Achieva 7T — marks a pivotal advancement for neuroimaging, enabling visualization of cortical layers, hippocampal subfields, and small vessel disease at sub-millimeter resolution. Ultra-high field systems require sophisticated radiofrequency engineering to address B1 field inhomogeneities and dielectric effects at higher frequencies. Concurrently, zero-echo time (ZTE) and pointwise encoding time reduction with radial acquisition (PETRA) sequences have enabled near-silent MRI by eliminating the abrupt gradient switching responsible for acoustic noise, dramatically improving patient comfort and enabling imaging of young children without sedation. These developments are comprehensively discussed in From 1.5T to 7T and Beyond: How MRI Technology and AI are Reshaping Diagnostics.

Portable and Low-Field MRI

The emergence of portable, low-field MRI systems represents a democratization of neuroimaging. The Hyperfine Swoop system (0.064T) operates at standard electrical outlets without RF shielding and can be deployed at the bedside in intensive care units, emergency departments, and resource-limited settings. Although spatial resolution and SNR are constrained at low field strengths, AI-enhanced reconstruction substantially closes the image quality gap for targeted clinical applications such as intracranial hemorrhage detection and hydrocephalus monitoring. These platforms are expanding access to diagnostic imaging in global health contexts where conventional high-field MRI infrastructure is cost-prohibitive.

Hybrid PET/MRI Systems

Simultaneous PET/MRI integration combines the metabolic and molecular sensitivity of positron emission tomography with the superior soft tissue contrast and functional capabilities of MRI — all within a single examination and without additional radiation exposure beyond the PET tracer. Systems such as the Siemens Biograph mMR and GE SIGNA PET/MR require avalanche photodiode (APD) or silicon photomultiplier (SiPM) detectors engineered to operate within high magnetic fields, representing a significant hardware design challenge. Clinical applications include oncologic staging, neurological research, and cardiac viability assessment. Attenuation correction — traditionally performed using CT in PET/CT — must instead rely on MR-based tissue segmentation, an active area of algorithmic development.

9. Key Takeaways and Tips for Biomedical Engineers

Engineering Principles to Master

A thorough command of MRI physics — including spin dynamics, k-space theory, pulse sequence design, gradient hardware, and RF coil engineering — is foundational for any biomedical engineer working in diagnostic imaging. Understanding signal-to-noise ratio optimization, parallel imaging acceleration (GRAPPA, SENSE), and the tradeoffs between spatial resolution, temporal resolution, and image contrast empowers engineers to evaluate system performance critically. Proficiency in image reconstruction mathematics, including Fourier transform theory and iterative algorithm design, is increasingly valuable given the shift toward AI-based pipelines. Engineers should also develop familiarity with cryogen management, superconducting magnet quench procedures, and RF shielding design. Exploring Top Skills Every Biomedical Engineer Should Master provides a broader framework for professional development in this field.

Regulatory and Standards Knowledge

Biomedical engineers involved in MRI system design, procurement, or quality assurance must be conversant with the primary regulatory and standards framework. IEC 60601-2-33 defines essential performance and safety requirements for MR equipment, encompassing static field limits, gradient dB/dt thresholds, SAR limits, and acoustic noise specifications. ASTM F2052 governs assessment of ferromagnetic qualities of passive implants, while the FDA’s guidance on MR-conditional labeling provides a pathway for implantable device manufacturers. ISO 13485 underpins quality management system requirements across the device lifecycle. Engineers should also track FDA 510(k) and PMA submissions for novel MRI technologies and understand post-market surveillance obligations. Familiarity with Key Organizations and Bodies in the Medical Device Field is essential for navigating the global regulatory landscape.

Career and Maintenance Tips

Clinical MRI engineers and MRI physicists play critical roles in equipment commissioning, acceptance testing, quality control (QC) programs, and preventive maintenance — including helium level monitoring, gradient calibration, and RF coil performance verification. Regular QC using standardized phantoms (ACR MRI phantom) ensures imaging performance remains within diagnostic thresholds. For engineers considering MRI-focused career pathways, certifications from the American Board of Magnetic Resonance Imaging (ABMRI) or the International Society for Magnetic Resonance in Medicine (ISMRM) strengthen professional credentials. Staying current with the rapidly evolving technology landscape — particularly AI reconstruction, ultra-high field systems, and new clinical applications — is essential for long-term career relevance. Biomedical engineers should also cultivate strong interdisciplinary communication skills to collaborate effectively with radiologists, physicists, and clinical staff throughout the device lifecycle.

References

1. Bushberg, J.T., Seibert, J.A., Leidholdt, E.M., & Boone, J.M. (2020). The Essential Physics of Medical Imaging (4th ed.). Lippincott Williams & Wilkins.
2. International Electrotechnical Commission. (2015). IEC 60601-2-33: Medical electrical equipment — Part 2-33: Particular requirements for the basic safety and essential performance of magnetic resonance equipment for medical diagnosis. https://www.iec.ch
3. ASTM International. (2015). ASTM F2052-15: Standard Test Method for Measurement of Magnetically Induced Displacement Force on Medical Devices in the Magnetic Resonance Environment. https://www.astm.org
4. U.S. Food and Drug Administration. (2021). Criteria for Significant Risk Investigations of Magnetic Resonance Diagnostic Devices — Guidance for Industry and FDA Staff. https://www.fda.gov
5. National Institute of Biomedical Imaging and Bioengineering (NIBIB). Magnetic Resonance Imaging (MRI). https://www.nibib.nih.gov
6. Siemens Healthineers. (2024). MAGNETOM Vida and MAGNETOM Prisma Product Pages. https://www.siemens-healthineers.com/magnetic-resonance-imaging
7. Hyperfine, Inc. (2023). Swoop Portable MR Imaging System — Clinical Overview. https://hyperfine.io
8. Lustig, M., Donoho, D., & Pauly, J.M. (2007). Sparse MRI: The application of compressed sensing for rapid MR imaging. Magnetic Resonance in Medicine, 58(6), 1182–1195. https://doi.org/10.1002/mrm.21391
9. Knoll, F., Hammernik, K., Zhang, C., Moeller, S., Pock, T., Sodickson, D.K., & Akcakaya, M. (2020). Assessment of data consistency for clinical evaluation of deep learning reconstruction in accelerated cardiac MRI. Radiology: Artificial Intelligence, 2(1), e190064. https://doi.org/10.1148/ryai.2019190064
10. Reiter, D.A., & Gold, G.E. (2022). Ultra-high field MRI: Clinical applications and safety considerations. Journal of Magnetic Resonance Imaging, 55(3), 681–700. https://doi.org/10.1002/jmri.27440