CompHealth Corner: August
Computational and digital health ideas evolve quickly. Our monthly research roundup highlights the latest publications — from our Center faculty and other leaders in the field. Here’s what you should be reading this month!
Physical Artificial Intelligence in Nursing: Robotics
Shaw, Ryan J., and Boyuan Chen. “Physical artificial intelligence in nursing: Robotics.” Nursing Outlook 73, no. 5 (2025): 102495.
Published in Nursing Outlook, this article explores how physical AI—the integration of advanced robotics with perception, reasoning, and action capabilities—is reshaping the future of care delivery.
Why you should read this:
The paper provides a cross-industry overview of robotics and zeroes in on their applications in skilled nursing. As robots evolve from logistical assistants to collaborative care partners, the authors highlight critical questions:
- How do we prepare nurses to work alongside robotic systems?
- Can robots truly augment care in meaningful, equitable ways?
- What ethical, regulatory, and workforce issues must be addressed?
Key takeaways:
- Demographic pressure is mounting. A growing aging population, declining birth rates, and a wave of nurse retirements are converging to strain our care infrastructure.
- Physical AI is accelerating robotic potential. Robots are already in hospitals for cleaning, fetching supplies, and basic support—but advances in physical AI could enable new roles in patient interaction, mobility assistance, and even virtual care.
- Nurses must help design the future. The paper argues that nurses should lead the development, deployment, and regulation of robotic systems—ensuring these technologies enhance human care rather than replace it.
Bottom line:
This isn’t science fiction—it’s already happening. And the future of robotic integration in nursing depends on proactive leadership from within the profession. As new models of care emerge, this paper calls for interdisciplinary collaboration, education reform, and ethical foresight.
Why we’re reading it:
At the intersection of technology and care, this is the kind of innovation that will shape tomorrow’s healthcare landscape. Whether you’re an engineer, clinician, educator, or policymaker, understanding the trajectory of robotics in nursing is critical to designing systems that support both equity and excellence.
Building Competency in Artificial Intelligence and Bias Mitigation for Nurse Scientists and Aligned Health Researchers
Cary Jr, Michael P., Siobahn D. Grady, Jacquelyn McMillian-Bohler, Sophia Bessias, Christina Silcox, Susan Silva, Vincent Guilamo-Ramos, Jonathan McCall, Jessica Sperling, and Benjamin A. Goldstein. “Building competency in artificial intelligence and bias mitigation for nurse scientists and aligned health researchers.” Nursing outlook 73, no. 3 (2025): 102395.
Published in Nursing Outlook, this paper tackles one of the most pressing questions in digital health: How can we ensure that AI/ML tools improve healthcare without reinforcing structural inequities?
Why you should read this:
AI is already influencing clinical decisions across the globe. But without deliberate intervention, these tools can replicate and even magnify disparities. This article introduces HUMAINE, a pioneering curriculum built to equip nurse scientists and healthcare researchers with the tools they need to:
- Detect and mitigate algorithmic bias
- Integrate social determinants of health (SDOH) into AI models
- Understand ethical and regulatory frameworks
- Collaborate across disciplines
- Promote equity and trust in AI-enabled care
Key takeaways:
- Bias in, bias out. Training diverse, equitable AI tools requires interdisciplinary literacy and intentional design.
- Nurses are frontline interpreters. As key caregivers, nurses will interface with AI daily—but few current programs train them to evaluate outputs critically.
- Education is the lever for change. HUMAINE’s curriculum spans technical AI skills, ethics, implementation science, and SDOH integration—all with the goal of preparing health researchers to lead responsibly.
As AI in health continues to evolve, programs like HUMAINE will be vital in training a future-ready workforce that can harness these tools for equity, not harm.
High-Performance Computing at a Crossroads
Deelman, Ewa, Jack Dongarra, Bruce Hendrickson, Amanda Randles, Daniel Reed, Edward Seidel, and Katherine Yelick. “High-performance computing at a crossroads.” Science 387, no. 6736 (2025): 829-831.
High-performance computing (HPC) has long been the silent engine powering advances in medicine, climate modeling, physics, and now AI. But the United States stands at a critical inflection point.
This policy piece, coauthored by leading voices in computing, is a call to action: the U.S. must adopt a long-term, comprehensive HPC strategy—or risk falling behind.
Why you should read this:
- HPC is no longer just for simulations. It’s now a core enabler of AI, bioinformatics, drug discovery, and national defense.
- The 2024 Nobel Prizes in physics, chemistry, and economics all highlighted the role of computing in scientific innovation.
- Other global powers (EU, China, Japan) are rapidly investing in HPC ecosystems with national plans, while the U.S. lacks a post-Exascale roadmap.
- Technical challenges are mounting: power consumption, chip fabrication costs, memory bottlenecks, and hardware mismatches with traditional modeling needs.
- And yet, opportunities abound: hybrid HPC+AI systems, custom silicon, codesign architectures, and deeper public-private collaboration.
Key takeaways:
- HPC is foundational to science, security, and economic competitiveness—but U.S. leadership is at risk.
- We need a 10-year national roadmap spanning hardware, software, workforce, and access, built with input from academia, industry, labs, and policymakers.
- Without action, we may lose not just leadership in HPC—but also the ability to lead in AI, precision health, and advanced science.
A “whole-nation” approach is no longer optional—it’s essential.
This paper is an urgent read for anyone working in science, tech policy, or computing.
Efficient, Automated Escape Routing for High-Density Pad Grids
Rachinskiy, Iakov, Dmitrii Rachinskii, and Jonathan Viventi. “Efficient, automated escape routing for high-density pad grids: deepest-exit-first algorithm for layer and wire count optimization.” Journal of Neural Engineering 22, no. 3 (2025): 036031.
As neural interfaces scale into the thousands of channels, designing compact, implantable devices becomes increasingly complex. One of the biggest bottlenecks? Routing all those tiny electrical traces out of an impossibly dense pad grid.
This new article introduces a deepest-exit-first algorithm that tackles this challenge head-on. Focused on thin-film (TF) neural interfaces, the method provides a routing solution for grids where wires can’t fit between adjacent pads—something current tools often can’t handle.
Why you should read this:
- Solves a real bottleneck in high-channel neural interface design, where trace routing can increase layer count and device size.
- Automates escape routing, easing design iteration and enabling tighter integration with wireless electronics and ASICs.
- Optimizes pad grids by intelligently removing non-essential pads—accepting that not every pad must be routed—improving flexibility and manufacturability.
Key takeaway:
This is more than a technical tweak—it’s a practical solution for the next generation of neural implants. As TF devices gain traction in research and clinical applications, efficient escape routing will be vital for maintaining implantability, reliability, and performance.
