Matt Rosen and Colleagues’ Journey Through the Secret Life of Plants


Gary Boas

In the waning months of 1979, the legendary Motown artist Stevie Wonder released an album called Stevie Wonder’s Journey Through “The Secret Life of Plants,” the soundtrack to the documentary film The Secret Life of Plants. Equal parts frustrating and strangely compelling, and notably using some of the most advanced music technology of the day, the album sought to shed light on the mysteries of the botanical world in part by probing its constituent parts. Wonder sang of seeds and leaves, and of plants’ electrical impulses, and in doing so hoped to reveal something about the nature of life.

Now, the Martinos Center’s Matt Rosen and colleagues have done him one better.

In a paper published online last week in the journal Geoderma, the researchers report a method for noninvasive imaging of root morphology and architecture in intact soils. The method uses low-field magnetic resonance imaging (LF-MRI) technology based on work Rosen and his group have done over the past decade, imaging humans with low-cost MRI technology, to help address previously insoluble puzzles of the botanical realm.

So much of plants’ lives transpire under the ground, where their roots snake through the soil seeking the nutrients they need to survive. This makes it difficult to study plants. Because soil is opaque to conventional imaging methods (as well as to the naked eye, of course), scientists have not been able to follow roots’ growth in a natural setting, where it can be impacted by changes in weather or soil nutrients or any number of other factors. And if they study plants in an artificial environment, where the roots are accessible, they lose many of the insights they would glean from following them in a natural one.

This problem proved so vexing that the U.S. Department of Energy’s Advanced Research Projects Agency – Energy (ARPA-E) decided to fund a project to develop a tool that could be used to understand the below-ground phenotype of plants – in particular, as a tool to help plant breeders develop plants that sequester more carbon from the air into their roots.

Several years ago, Rosen, Assistant Professor of Radiology, Director of the Low-field MRI and Hyperpolarized Media Laboratory, and Co-Director of the Center for Machine Learning, both at the Martinos Center, received a call from ARPA-E wanting to know if he would be interested in such a project. “They knew about my work as a physicist doing unconventional MRI and asked me to attend a meeting where I met the soil scientists who would become my collaborators,” he says. Rosen was tasked with leading the Sensor team, the half of the project that developed the actual imaging hardware that would go into the ground. The lead PI on the project would be Texas A&M soil scientist Cristine Morgan – “the best soil scientist in the world,” Rosen says.

Through this collaboration, Rosen and colleagues developed advanced MRI technology operating at 47 mT (that is, 64 times less than conventional human MRI at 3 Tesla) for imaging of energy sorghum root morphology and architecture in agricultural soil. The use of very weak magnetic fields both allowed for construction of a portable MRI device for field applications and reduced the distortion caused by naturally occurring magnetic material found in the soil. This technology was bolstered by the use of a data-driven machine learning reconstruction approach – Automated Transform by Manifold Approximation (AUTOMAP) – developed in Rosen’s lab. The application of AUTOMAP showed more than a two-fold improvement in signal-to-noise ratio on average, further enhancing the quality of the final images.

In the Geoderma paper, the investigators describe a demonstration of the efficacy of the technology for imaging roots in soil. The next step, they say, is to optimize it for deployment in the field, in part by scaling up the size of the magnet to generate higher-resolution images in a shorter amount of time.

For Rosen, one of the highlights of the project has been working with so many outstanding researchers across the country. The Sensor team includes members of Rosen’s lab at the Martinos Center, a group at National Institute of Standards and Technology (NIST) in Boulder, Colo., and the team at ABQMR in New Mexico. “ABQMR is Dr. Eiichi Fukushima’s company, and Eiichi is a world-renowned NMR pioneer,” Rosen says. “He is also a scientific hero of mine, so to have the opportunity to put together my ‘dream team’ with all these great folks was an opportunity not to miss.”

He adds, “the A&M team – including Cody Bagnall, the graduate student who took all the data in the hot sun in College Station – taught us everything we could learn about soils, roots, doing things in the field far from anything, and just dedication and hard work.”

Equally gratifying was the opportunity to perform a new kind of translational research.

“It is rewarding to leverage all the work we’ve done in my lab over the past ten-plus years making MRI more accessible for human imaging to a project in a different domain: MRI of plant roots in the ground!” he says. “This will lead to new understandings for plant breeders on how plant genotypes lead to specific desirable root phenotypes, and hopefully provide more soil security and reduce atmospheric carbon.”