With chronic back pain, discomfort and uncertainty often outlast solutions. Through a five-year, $3 million National Institutes of Health grant, C.-Y. Huang, a University of Miami College of Engineering professor, is working to validate a model that could help diagnose and treat spinal conditions.
As a biomedical engineer, Huang is leading a multidisciplinary team to identify the biological drivers of disc degeneration and develop artificial intelligence tools that could help physicians forecast how a patient's spine will change over time. Rather than focusing only on what damaged discs look like on a scan, Huang's research examines what is happening inside the disc at the cellular level.
Degeneration of the intervertebral discs, the cushions between the bones of the spine, is often associated with lower back pain. Much of the existing research has focused on inflammation and tissue breakdown. Huang believes another factor may play a larger role: whether disc cells are getting enough fuel to survive and function.
"Spinal discs have no blood vessels," Huang said. "Cells inside the disc rely on nutrients diffusing in from surrounding tissue. If they don't get enough nutrition, they cannot produce the matrix."
In this context, nutrition refers to the supply of glucose and other nutrients that must travel through surrounding tissue to reach disc cells. This is their main energy source. That energy allows them to produce molecules that attract water and keep the disc cushioned and flexible. When the nutrient supply drops, the cells produce less of those molecules and the disc degenerates.
"In my opinion, the major factor causing disc degeneration is nutrition," Huang said. "Inflammation is there, but when you look at the concentration of inflammatory cytokines in blood and tissue, it is actually very low."
Huang and longtime collaborator Weiyong Gu, a professor of mechanical and aerospace engineering at the University of Miami, have spent years developing a mathematical model of the human intervertebral disc. The model simulates how nutrients move through tissue and how cells respond under different conditions, including aging and mechanical stress.
"We can put any factor into the model and run simulations for different situations," Huang said. "We can isolate one factor and see how it causes disc degeneration."
Until now, much of that work has not been patient-specific. With the new NIH funding, the team will validate the model using human data, including MRI and PET-CT scans and disc tissue collected during surgery with assistance from Dr. Timur Urakov, a neurosurgeon at UHealth.
"Imaging tells you what the disc looks like," Huang said. "The model prediction and simulation tell you what the cells are actually doing."
Once validated, the model will generate large amounts of simulated data to train an AI system in collaboration with co-principal investigator John Martin at Rush University.
Instead of running time-consuming simulations, physicians could eventually use the AI tool to analyze a patient's scans and estimate how the disc is likely to progress or respond to treatment. While the research is still in its early stages, Huang said validating the model is the critical first step in giving physicians a more precise tool to treat patients.
