The University of Michigan is at the forefront of utilizing cutting-edge technology to solve complex problems that were once deemed insurmountable. By intertwining high-performance computing (HPC) and artificial intelligence (AI), researchers are making strides in various fields, leading to groundbreaking discoveries and advancements.
The innovative strategy in play at U-M combines robust hardware with sophisticated computational algorithms. This synergy is essential in accelerating scientific computing, enabling the analysis of intricate data and simulations. According to Karthik Duraisamy, Arthur B. Modine Professor of Engineering and Director of the Michigan Institute for Computational Discovery and Engineering, “When computational algorithms, infrastructure and AI advance together, they dramatically expand what we can model, predict, design and discover.” The university’s commitment to this approach is evident in its partnership with Los Alamos National Laboratory, aiming to establish a supercomputing facility in Washtenaw County.
AI Detects Hidden Cancer in Surgery
At Michigan Medicine, Todd Hollon is leveraging AI to enhance the diagnosis and treatment of gliomas, which are the most prevalent type of adult brain tumors. As the Joseph R. Novello M.D. and Alfredo Quinones-Hinojosa M.D. Research Professor of Neurosurgery, Hollon is pioneering the integration of AI in surgical procedures.
Gliomas are notoriously difficult to treat due to their tendency to integrate with healthy brain tissue. Traditional methods, such as MRI scans or fluorescent dyes, often fail to detect all cancer cells. A new AI system developed at U-M addresses this issue by identifying residual tumor cells during surgery, delivering diagnostic feedback within approximately 10 seconds. This innovation reduces the rate of undetected tumors from 24% with conventional techniques to merely 4%, significantly enhancing patient outcomes.
“Our AI system improves surgical accuracy by up to four times compared to standard methods,” Hollon stated, highlighting the potential of this technology to transform the treatment of various cancers and significantly improve patient survival rates.
Innovations in Regenerative Medicine
Associate Professor Indika Rajapakse is spearheading efforts to reprogram skin cells into bone marrow using AI and HPC, a groundbreaking approach that could revolutionize treatments following chemotherapy. Bone marrow is vital for producing cells that combat infections and facilitate tissue healing.
By activating specific genes, Rajapakse’s team can convert a patient’s skin cells into bone marrow cells, dramatically reducing the risk of immune rejection. “We call this work the treatment for the treatment,” Rajapakse noted, emphasizing the potential of these technologies to tailor recovery processes to individual patients.
This approach promises wide-ranging applications, from regenerating heart muscle cells post-heart attack to developing insulin-producing cells for diabetes patients. Computational modeling, unconstrained by time or safety concerns, allows researchers to simulate numerous scenarios, uncovering insights that traditional methods cannot.
Predicting and Mitigating Risks
U-M researchers are also working on predicting space weather as the sun nears its activity cycle peak. Geomagnetic storms can disrupt satellite operations, power grids, and other critical systems. Using HPC and AI, the team models solar wind patterns, aiming to provide early warnings to protect essential infrastructure.
Moreover, U-M engineers are exploring the potential of lightweight, durable electric batteries for use in electric vertical takeoff and landing vehicles (eVTOLs). These vehicles could revolutionize emergency response by delivering medical supplies or personnel quickly to otherwise inaccessible locations.
The university’s use of HPC and AI extends to assessing building responses to severe events. Simulated scenarios help identify vulnerabilities that could lead to structural collapse, guiding improvements in building safety.
A Vision for Societal Benefit
U-M’s broader strategy focuses on leveraging HPC and AI for societal advancement. Arthur Lupia, vice president for research and innovation, stated, “AI is advancing at a pace unlike any technology we’ve seen before.” While challenges exist, the university is committed to ensuring these technologies serve the greater good, setting ethical standards and influencing policy to harness AI’s full potential.
