Cracking Carpal Dynamics: A Riddle, Wrapped in a Mystery, Inside a Wrist

Next time you shake hands or wave hello in a friendly greeting, consider your wrist, the small fulcrum between your forearm and your hand that makes these gestures possible. The wrist plays an essential role in the manual dexterity so characteristic of our species, yet surprisingly little is known about what happens inside it during even the simplest movements. Now, scientists at NYU Langone Health and the Center for Advanced Imaging Innovation and Research are combining magnetic resonance imaging (MRI), flexible radiofrequency (RF) hardware, machine learning, and 3D modeling techniques to shed light on this important bit of human anatomy.

“Decades and decades have been dedicated to the elucidation of the mystery of the wrist,” said Catherine Petchprapa, MD, associate professor of radiology at NYU Grossman School of Medicine, whose long-standing interest in this joint has helped motivate the new research. The project has the potential to uncover new information about the kinematics of the wrist and pathologies like wrist instability, which can manifest as a snapping sensation or misalignment. Carpal instability can be painful and, if left untreated, may lead to osteoarthritis.

An Elusive Snap

“That happens very quickly and spontaneously,” said Dr. Petchprapa of a typical wrist pop. “So there’s a need for not only imaging the joint in motion but imaging the joint fast enough to catch these very transient abnormalities.”

Current in vivo imaging methods, however, are more akin to stills than video and unable to reliably capture such fleeting events, said Dr. Petchprapa. And the methods that do visualize motion don’t provide three-dimensional data. This is due to a fundamental technical tradeoff between imaging speed and the amount of acquired information: capturing a three-dimensional volume takes longer than obtaining a projection or a “slice”—too long to render fine details of small, fast-moving anatomy.

“Some signs of carpal instability can only be seen on dynamic images,“ said Riccardo Lattanzi, PhD, professor of radiology at NYU Grossman and scientist at the Center for Advanced Imaging Innovation and Research, who is the principal investigator on the project. People with symptoms are often referred for fluoroscopy, a dynamic X-ray imaging technique. However, “fluoroscopy is a 2D projection, so you cannot see through-plane motion,” said Dr. Lattanzi. “And it’s also qualitative,” meaning that the evaluation relies on the radiologist’s judgment, but “there’s no quantitative information.”

Imaging speed isn’t the only obstacle to understanding wrist instability and, more fundamentally, to understanding healthy wrist dynamics. Dr. Petchprapa explained that X-ray methods, whether static (radiography), dynamic (fluoroscopy), or three-dimensional (computed tomography), are “very good at capturing the pictures of the bones but cannot tell you anything about the structures surrounding them—the muscles, the tendons, the nerves.” In addition, these modalities expose people to ionizing radiation, posing incremental risk with every scan and hindering the kind of repeat examination required in research.

Scientific efforts at measuring the biomechanics of the wrist have so far had to make do with these limitations. To circumvent worries about excessive radiation, some studies have resorted to the use of cadaveric arms animated by mechanical rigs intended to simulate natural movement. But ex vivo approaches can only teach us so much. “It isn’t in a living person where muscles contract to move and stabilize the joint and nerves fire and the range of motion is limited by pain,” said Dr. Petchprapa.

Although MRI comes with the same speed-versus-coverage trade-offs as do X-ray modalities, it has two key advantages: it can provide a detailed view of soft tissues and it does not impart harmful radiation.

Dr. Petchprapa’s fascination with the wrist began in the early 2000s, after her fellowship. Around 2010, she began working with MRI technologists at NYU Langone to explore imaging wrist motion at natural speed but eventually put the idea on hold as it became clear that the project required greater resources. In 2017, she mentioned the research to Dr. Lattanzi, with whom she was working on an investigation in hip imaging. “He was immediately interested,” she said.

Dr. Lattanzi, who holds a joint appointment at NYU Langone and NYU Tandon School of Engineering, involved colleagues at NYU Tandon who brought expertise in motion analytics. The ensuing collaboration led in 2022 to an award from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) through the R21 mechanism that, according to the institute, supports “innovative, ground-breaking projects with potential for significant impact.”

Eight Tiny Bones, One Big Puzzle

The wrist comprises eight carpal bones arranged in two rows and linked by a network of ligaments and muscles. The variety of shapes in this bony octet is evident in the carpals’ names: scaphoid (boat-shaped), lunate (moon-shaped), triquetrum (triangular), pisiform (pea-shaped), trapezium and trapezoid (table-shaped), capitate (head-shaped), and hamate (hooked).

The carpals are bookended on the hand side by the metacarpals and on the forearm side by the radius and the ulna. The ensemble, in which each bone can shift against its neighbors, forms an intricate and compact system that produces a wide range of motion. For now, the investigators have narrowed their focus to a simple wrist movement called radial-ulnar deviation—a side-to-side tipping of the hand toward the thumb, then toward the pinky, and back again. 

Using flexible high-impedance coil technology pioneered at the Center for Advanced Imaging Innovation and Research in 2018, the team created an RF coil that resembles a compression sports brace. Unlike traditional rigid RF coils, a high-impedance design can flex with the anatomy of interest, almost like a garment.

“You just wrap it,” said Ruoxun Zi, MPhil, doctoral candidate in biomedical imaging and technology at NYU Grossman and member of the research team, during a recent demonstration of the imaging hardware. The researchers also designed a 3D-printed platform that helps standardize the movements performed in the scanner. “The idea is that people only move the wrist and keep the arm in the same position,” said Zi. With her hand clad in the coil, she strapped her forearm onto the platform and gently gripped a hinged pad. By swiveling the pad side to side, she alternated between deviating her hand in the direction of the radius and the ulna, mimicking what the study participants do in an MRI scanner.

The researchers image this motion with a “LiveView” sequence that captures thin slices approximately 20 times per second, a speed similar to the frame rate of conventional video. Markers built into the hinged platform allow the team to use a machine learning algorithm to automatically align the reconstructed images, stack them, and create synthetic four-dimensional MRI volumes (three spatial dimensions plus time).

LiveView was first developed in the late 2000s at Max Planck Institute for Biophysical Chemistry by Tobias Block, PhD (then doctoral candidate) and colleagues. In the early 2010s, after joining NYU Langone, where he is now an associate professor of radiology, Dr. Block collaborated with Dr. Petchprapa on dynamic wrist MRI, but “at the time, we only had the ability to image individual slices,” he wrote in an email. A decade later, the addition of platform markers and machine learning has enabled “fusing sequentially-scanned slices into a 4D dataset, [and] now allows overcoming that earlier limitation.”

Next come the post-processing and computer modeling. Batool Abbas, PhD, postdoctoral fellow at NYU Langone, has developed a pipeline that starts with what she calls a template: “essentially, a statistical descriptor that represents the wrist in a neutral position created by averaging all the different poses the wrist takes over the course of the MRI sequence,” she said. The template lays the groundwork for segmentation. Dr. Abbas then uses a semi-automated process to segment carpal bones in the template image before back-propagating that segmentation to each frame in the sequence—a procedure that results in a 3D model of the wrist bones without frame-by-frame manual labor. 

The segmentation algorithm “is nonlinear and inevitably introduces small inconsistencies,” Dr. Abbas said. “Because you want the model to be true to how the bone behaves, we add a rigidity constraint to make sure that in every frame the bone is exactly the same shape. So, when we eventually calculate trajectories, we want to ensure that any changes are due to the bones’ motion.”

Those trajectories—extracted from each modeled bone’s center of mass and graphed along X, Y, and Z axes—are the first draft of an emerging method for quantifying carpal motion.

The idea is that a kink in such a trace could identify abnormal movement, like a snap, while the accompanying MRI would provide a view of both the bones and the surrounding soft tissues.

Beyond Instability

Dr. Lattanzi called the results promising and noted that the investigation is still in early stages. “We still need to show that any change in the trajectory is due to pathology,” he said. The team has shared its work in presentations at the 2023 and 2024 meetings of the International Society for Magnetic Resonance in Medicine and is currently conducting variability testing. The next likely step is a controlled study. If successful, it “would be the first time that we can reveal normal and pathological kinematics in the wrist,” said Dr. Petchprapa.

Quantification of carpal kinematics would carry implications for more than just medicine. Anatomists and evolutionary anthropologists have identified the wrist as critical to the functional adaptations that distinguish humans from other primates. These adaptations include precision grips that allow us to make and use tools, throw projectiles, and swing clubs—motions that continue to define us even as we manipulate objects far more refined than the sticks and stones handled by our ancestors.

“It starts as a clinical question but leads into a more fundamental inquiry,” said Dr. Petchprapa about the group’s research. “It’s really interesting to look at [the wrist] as an adaptation that was extremely successful for our species.”

Research reported in this story was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health (NIH) under award number R21AR080325. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

This story was published on August 9, 2024, and updated on August 15 with additional information about the LiveView sequence and a list of related publications.