Molecules, Math and MRI

The renowned British poet Elizabeth Barrett Browning composed one of the most famous love sonnets of all times: How do I love thee? Let me count the ways. It is a declaration of the love she had for her husband, Robert Barrett. Had she been a 21st century radiologist with an equal gift for words, she might have written a poem praising the glories of magnetic resonance imaging (MRI). It might have begun, “What are thy wonders? Let me count the waves.”

MRI has rightly been called a wonder of modern physics. In order to generate images of bodily structures, it relies on radiofrequency (RF) waves interacting with a powerful magnetic field to act upon water molecules in tissue. When pulsed RF waves are transmitted into the magnetic field, they stimulate the protons in the molecules’ atoms to spin. Between pulses, the spin quickly relaxes. During the entire process, energy is released as additional radio waves that are picked up as signals by a receiver. These signals are then translated into images.

One of the beauties of MRI is the ability to adjust the signal intensity by varying imaging sequences that “govern” the spin and the timing of the relaxation. In physics terms, a pair of time constants called T1 and T2 describes this dual process. Each sequence “interrogates” the chemical environment of the water molecules within it, revealing particular tissue properties. The more sequences used, the more information gained. It’s a little like a reporter interviewing as many different witnesses at an accident scene asking who, what, where, when, why and how in order to piece together the whole picture.

What MRI can and can’t do

MRI has come a long way since it was invented by Dr. Robert Damadian in 1972. Today’s MRI creates detailed portraits of structures in the body. It is safe, noninvasive, and has no exposure to radiation as occurs with x-rays and CT scans. It produces high resolution, 3D images of anatomy; even more importantly, it gives information on specific tissue characteristics depending on the pulse sequences. Improvements in MRI allow imaging not only of organs, bones, cartilage, and tumors or other abnormalities in those structures, but it can also be used to analyze even very small blood vessels and nerves. Not only is it used for diagnosis, it can also show whether treatments are having the desired healing effects on targeted areas.

What MRI cannot do—at least, not yet— is depict underlying disease processes at the molecular level in the body. At least a decade of research has gone into expanding MRI’s abilities to meet the challenge of real-time biomolecular (molecules in living tissue) scanning. As author John D. MacKenzie puts it,

The goal is to reveal the early underlying biochemical and genetic events responsible for disease rather than indirect and late changes (eg, altered blood flow or tumor size) as seen with most current clinical diagnostic imaging modalities. Direct imaging of events fundamental to disease processes with molecular imaging should ultimately translate into better patient care through earlier and more specific detection and intervention.[i]

Biomolecular MRI: How math can help

How have researchers tried to accomplish this? Most of the approaches so far have involved the idea of adding some type of molecular agent that would be taken up by specific biomolecules that would “light up” in an MRI scan. In principle, a new type of tailored contrast agent could be injected into a blood vessel for distribution to targeted receptors on cells. Some contrast agents would be designed to target cancer cells; others to inflammation; others to support gene or cell-based therapies, etc. However, it will take much time to develop and test such agents to the point where they can be commonly used.

A new approach using mathematical probability takes advantage of the data already present in any given MRI scan. A March 22, 2019 article in Science Magazine reports a new statistical technique that extracts the data without the need for a contrast agent in a patient’s body. The key lies in disentangling pairs of the T1 and T2 time constants, “…one pair for the water surrounding a biomolecule of interest and the other for the water farther away. Disentangling those two sets is devilishly difficult, akin to listening to the same note played simultaneously on two xylophones and telling which note faded faster.”[ii]

Two physicists, Richard Spencer and Mustapha Bouhrara created a statistical approach that results in a probability distribution for each time constant. Their initial efforts have so far required a time-heavy computation but it will get more efficient with development. Already, applying their method to brain MRI scans, Spencer and Bouhrara reported that “…adults with mild cognitive impairment and Alzheimer’s disease have less myelin [the insulation that facilitates nerve transmission between neurons] in their brains…”[iii]

It’s exciting, but it’s not quite ready for prime time. It will take some years of gathering images from healthy people and running them through this math approach in order to develop baseline information. On the other hand, the use of molecular contrast agents also has a long way to go. As with all developmental science, the race is on to see whether math or a molecular contrast agent gets to benefit humanity first.

NOTE: This content is solely for purposes of information and does not substitute for diagnostic or medical advice. Talk to your doctor if you are experiencing pelvic pain, or have any other health concerns or questions of a personal medical nature.


[i] MacKenzie, John. Molecular Magnetic Resonance Imaging. Appl Radiol. Jan 21, 2005. https://appliedradiology.com/articles/molecular-magnetic-resonance-imaging

[ii] Cho, Adrian. Clever math enables MRI to map biomolecules. Science 363 (6433), 1263.

[iii] Ibid.

This site uses cookies to analyze traffic and user behavior, protect your privacy, and provide you with the best user experience. Learn more.
close-image

How can we help?

Contact us to discuss your prostate health and plan your path to wellness.

close-link