Why doctors removed the word 'nuclear' from MRI scans
When you undergo an MRI, you aren't being exposed to X-rays or radiation. While the technology relies on the magnetic properties of atomic nuclei, the name was changed to avoid the negative associations people have with nuclear energy.
Originally known as Nuclear Magnetic Resonance Imaging (NMRI), the term was shortened to MRI to distance the procedure from the stigma of nuclear physics. The technology actually functions by using powerful magnetic fields and radio waves to interact with the hydrogen atoms naturally abundant in our bodies, particularly within water and fat. By manipulating these atoms, clinicians can map the precise location of different tissues.
The process is remarkably complex. An MRI scanner uses a main magnet to polarize the body's protons, while gradient coils create localized variations in the magnetic field. When radiofrequency pulses are applied, the nuclear spin of these protons resonates, emitting a signal that is detected by antennas. This signal is then processed by computers to construct detailed digital images of internal anatomy.
The utility of the scan lies in 'relaxation'—the rate at which excited atoms return to their equilibrium state. T1-weighted images are excellent for identifying fatty tissue or anatomical structures, while T2-weighted images are vital for detecting inflammation or edema. While highly effective for visualizing soft tissues like the brain or abdomen, the technology has limitations. The loud, rhythmic clicking of the machine and the confined space can be distressing for claustrophobic patients, and certain metallic implants can pose significant safety risks.
Modern advancements are pushing these boundaries even further. While most clinical systems operate at 1.5 or 3 Tesla, researchers are exploring much higher fields, such as 14 Tesla. Simultaneously, the integration of artificial intelligence is revolutionizing the field; projects like the 2018 fastMRI collaboration between Meta AI and NYU Langone Health have demonstrated that deep learning can accelerate brain and knee scans by up to four times without sacrificing diagnostic accuracy.
Source: Magnetic resonance imaging
