As space colonization efforts advance, WVU scientists are developing AI-powered computer models to treat and prevent the physical challenges astronauts face due to extended weightlessness.
A study led by Valeriya Gritsenko and Sergiy Yakovenko, associate professors in the departments of Human Performance and Neuroscience in the WVU School of Medicine and WVU Rockefeller Neuroscience Institute, responds to the growing likelihood that astronauts will be spending longer periods in microgravity environments like space stations, the moon or rocket ships bound for Mars.
For each astronaut, Gritsenko and Yakovenko’s technology will be able to create a unique “digital twin” — a computer model capturing the relationships between that person’s movements and muscle activity.
The models will show how each astronaut adapts to weightlessness and identify what that specific individual needs to do to counteract the well-known hazards of microgravity, including muscle loss, declining bone density, and the vision and neurological changes that emerge when gravity disappears.
“Currently, each astronaut requires a very large Earth-based team that looks at his or her vitals and modifies exercise plans or otherwise intervenes,” Yakovenko said. “As we travel farther away from Earth, that kind of support will not be available, so we are researching alternatives.
“When astronauts return to Earth, they say what they miss the most in space is the ability to walk,” he added. “Even though they exercise a lot, they cannot maintain the correct coordination in microgravity. So, if we want to travel as far as Mars, the astronauts’ coordination will be disrupted in a profound way. Right now, they might make it to Mars, but by the time they set foot on the planet, they would probably fall down. We don’t want that, so we are developing a non-invasive way of observing astronauts as they go about their business.”
To build the digital twin models, the team will study human volunteers in a carefully controlled virtual reality environment, asking them to complete tasks like reaching for objects or walking on a treadmill. Motion-capture systems and wearable sensors will track their movements and record the underlying neural drive to the muscles.
“Our virtual reality software is similar to the artificial physics engines used for gaming or movies, which simulate the movement of virtual characters in a way that looks natural,” Gritsenko said. “Given factors like a person’s size, weight and muscle action, we can run simulations to see what kind of forces people would need to make certain movements, and that predicts what astronauts would experience in orbit or in a zero-gravity environment.”
Their AI models will monitor and analyze each astronaut’s structured exercise routine before the spacecraft launch, throughout the mission, and after returning to Earth. By continually tracking how the astronauts move, the system can spot subtle early signs of trouble long before they become real problems.
“The model can then tell the astronaut, ‘Exercise more, or use heavier weights or else you might be at risk of some muscle loss,’” Gritsenko said. “When they come back to Earth, the model would also have a reliable estimate of how much they have de-conditioned in orbit and could develop a routine to help them cope with problems like balance issues or ‘orthostatic intolerance,’ when they lose consciousness because blood is not pushed into their heads quickly enough when they stand up.”
The project is supported by a $750,000 grant from NASA to the NASA WV Space Grant Consortium and NASA WV Established Program to Stimulate Competitive Research, or EPSCoR.
“NASA WV EPSCoR is proud to support research led by Drs. Gritsenko and Yakovenko that pushes the boundaries of human space exploration while creating opportunities for West Virginia students to contribute to groundbreaking discoveries,” Candy Cordwell, assistant director of the NASA WV Space Grant Consortium, said. “Projects like this not only advance NASA’s mission but also attract federal investment, drive high-tech innovation, and prepare our students for careers in aerospace, AI and biomedical fields. These efforts position West Virginia as a key contributor to the growing space economy and help fuel long-term economic growth in the state.”
WVU students from within the School of Medicine and other disciplines that range from computer science to engineering are contributing to the research. Gritsenko said the confluence of the different backgrounds the student researchers bring is “where the magic happens.” The research team currently includes graduate students Nishat Tasnim Koli, Seyed Rasoul Hosseini and Shawn Shultz, as well as undergraduate students Natalie Esposito, Ian Copenhaver, Abigail Stewart and Maya McKendall.
In addition to space travel, the AI models can be adapted for rural telemedicine and training surgeons to conduct minimally invasive surgeries.
“People can get more out of telehealth if we use these digital twin tools for early detection of concerns like motor deficits, balance problems or even early neurodevelopmental delays,” Gritsenko said. “There are lots of parallels here between spaceflight de-conditioning and muscle and neural motor control in people who are very sedentary — for instance, those who are older or pregnant people who have been on extended bed rest.”