Partial differential equations function as mathematical expressions interlacing multiple variables, capturing how physical quantities evolve over space and time. Traditionally, scientists and engineers rely on immense numerical simulations that require powerful electronic processors and substantial energy consumption to approximate PDE solutions. Existing computational methods—whether through direct numerical approaches or electronic neural networks—struggle to balance precision, speed, and scalability, especially when tackling large-scale scientific problems with intricate boundary conditions.

The University of Utah team, led by Assistant Professor Weilu Gao and Ph.D. candidate Ruiyang Chen, has confronted this challenge by reimagining PDE computation through an optical framework. Their ONE device fundamentally departs from digital paradigms by encoding PDE variables as distinct attributes of light waves, including intensity, phase, and polarization. The light, modulated to represent an initial state, traverses a sequence of custom-designed diffractive optical elements and matrix multipliers, dynamically altering its properties to embody the PDE solution as it emerges.

At the core of this approach lies a sophisticated interplay between photonics and neural network principles. Neural networks traditionally process inputs by propagating signals through layered computational nodes, each applying weighted transformations to steer the network towards a final output. The optical neural engine replicates this process in the physical domain, where photons carry information across diffractive surfaces that perform matrix multiplications and function approximations akin to digital neurons. This physical embodiment sidesteps the bottlenecks of electronic signal processing, leveraging the inherent parallelism and speed of light propagation.

One of the key advantages of the ONE system is its remarkable acceleration of the machine learning workflow central to solving PDEs. Electronic neural networks, while powerful, consume considerable energy and take longer to converge on solutions. The optical method requires significantly less power, as photons naturally propagate without resistive losses inherent in electronic circuits, and computation occurs at the speed of light within the optical hardware. According to the research team, this translates to dramatic reductions in both execution time and operational costs for complex PDE problems.

The practical applications demonstrated by the ONE device underscore its versatility. The researchers tested the system on diverse PDEs including the Darcy flow equation, which models fluid movement through porous media; the magnetostatic Poisson’s equation relevant to demagnetization studies; and the Navier-Stokes equation governing incompressible fluids. In each case, the optical neural engine successfully learned mappings between input parameters and output physical quantities, predicting solutions without resorting to traditional experimental or computational heavy lifting.

As Professor Gao explains, “The Darcy flow equation, for instance, depicts how fluids navigate through materials filled with tiny pores. Our optical neural engine effectively internalizes the relationship between permeability and pressure within that medium, allowing it to anticipate fluid behavior purely from encoded light signals.” This capability promises transformative impacts in fields such as geology, environmental engineering, and even microchip design, where modeling complex interactions swiftly and accurately is paramount.

Beyond performance, the ONE’s optical paradigm introduces new opportunities for real-time scientific computations that were previously impractical. Situations demanding near-instantaneous PDE solutions—such as adaptive control systems, weather prediction, or biomedical imaging—could greatly benefit from integration with optical neural components. The research opens a pathway toward photonic computing platforms capable of handling dynamic, high-dimensional data with efficiency unattainable by existing electronic processors.

The study, published in Nature Communications, not only presents the hardware innovation but also details the computational modeling substantiating the ONE’s functionality. Through extensive simulation and experimental verification, the team demonstrated that the optical components perform matrix multiplications and nonlinear transformations integral to neural network operation with high fidelity. This foundational work paves the way for scaling the optical neural engine to tackle more complex PDE systems and potentially coupling it with electronic components for hybrid computational architectures.

Supporting the advancement are co-authors from both the University of Utah and Lawrence Berkeley National Laboratory, including Zhi Yao, Andrew Nonaka, Minhan Lou, and Jichao Fan, alongside collaborators from the University of Maryland. Their interdisciplinary effort embodies a fusion of electrical engineering, applied physics, and computational mathematics, highlighting the multi-faceted nature of contemporary scientific innovation.

Funding for this research was generously provided by the National Science Foundation, the U.S. Department of Energy, and institutional start-up grants, underscoring the strategic importance of advancing computational methodologies. As the global demand for computational power surges amid challenges in energy consumption and carbon footprint, technologies like the optical neural engine represent a hopeful stride toward greener and faster scientific computing.

In the broader context of machine learning and artificial intelligence, this optical neural engine signifies a provocative shift. While electronic AI chips have evolved rapidly, their reliance on semiconductor transistors imposes limits on processing density and energy efficiency. Photonic-based computation, exemplified by the ONE, offers an alternative frontier where light’s wave properties can perform analog computations inherently suited for complex mathematical operations, potentially leading to a new class of ultrafast, low-power AI hardware.

Ultimately, the University of Utah’s optical neural engine presents a compelling case for integrating optics and machine learning to solve some of science’s intractable problems. By transforming PDEs from abstract equations into dynamically evolving optical signals, the researchers have charted a course toward accelerated discovery and innovation across physics, engineering, and beyond. As this technology matures, it promises to redefine our computational toolkit, enabling scientists to simulate, predict, and understand the complexities of the natural world with a speed and scale never before possible.

Subject of Research: Partial differential equations (PDEs) solved through optical neural networks

Article Title: Optical neural engine for solving scientific partial differential equations

News Publication Date: 17-May-2025

Web References:
https://www.nature.com/articles/s41467-025-59847-3

References:
Gao, W., Chen, R., Yao, Z., Nonaka, A., Lou, M., Fan, J., Yu, C. (2025). Optical neural engine for solving scientific partial differential equations. Nature Communications.

Image Credits: Gao lab, University of Utah

Keywords: Partial differential equations, computational physics, optical neural networks, photonic computing, machine learning, diffractive optics, fluid dynamics, neural engineering

Traditional medical systems have long struggled with accurately predicting long-term health conditions. Patients who might develop significant health issues, such as depression or hypertension, are often overlooked, resulting in delayed intervention and treatment. Current methodologies achieve an identification accuracy of only around 50% to 75%. In contrast, RiskPath utilizes advanced time-series AI algorithms that have demonstrated an unprecedented accuracy rate of between 85% and 99%. This enhancement is attributed to the system’s ability to analyze extensive datasets collected over years, deciphering intricate patterns that indicate an individual’s risk profile for developing chronic diseases.

The implications of this technology are especially pertinent considering that chronic progressive diseases are responsible for over ninety percent of healthcare expenditures and mortality rates worldwide. Dr. Nina de Lacy, an assistant professor at the University of Utah Health and the study’s lead author, emphasizes the critical importance of early identification of high-risk individuals. By recognizing and analyzing which risk factors are most influential at various stages of life, healthcare professionals can craft tailored preventative strategies that address specific needs. This shift in focus from reactive to proactive healthcare is vital for improving patient outcomes.

RiskPath’s efficacy has been validated through extensive research across three large-scale longitudinal studies involving thousands of participants. Within these studies, the researchers successfully predicted a range of eight conditions, including anxiety, ADHD, and metabolic syndrome. The predictive ability of RiskPath not only enhances our understanding of disease development but also allows for a more nuanced view of how different risk factors can evolve in importance as individuals age. For instance, the research illustrated how factors like screen time and cognitive functioning can significantly impact the risk for ADHD as children transition toward adolescence.

Furthermore, RiskPath provides a streamlined risk assessment framework. While it possesses the capacity to analyze hundreds of health variables, the research unveiled that most conditions can still be accurately predicted using only a select set of ten key indicators. Such efficiency helps facilitate the application of RiskPath in clinical environments, as fewer data points make it easier for healthcare providers to implement this innovative model without overwhelming complexity.

Visualizations generated by RiskPath further add to its advantages, offering intuitive representations of an individual’s risk contributions over various life stages. By elucidating which periods contribute most significantly to the risk of disease, healthcare providers can discern optimal times to intervene, allowing for targeted preventive measures that could potentially alter the trajectory of health for at-risk populations.

Looking ahead, the team behind RiskPath is contemplating the integration of this technology into existing clinical decision support systems. By embedding RiskPath into preventive healthcare programs, they can enhance the toolkit’s utility for mental health practitioners and other healthcare providers. The ongoing exploration into the neural basis of mental illnesses will also play a pivotal role in refining this tool and expanding its applicability to additional disease domains and diverse demographic groups.

The potential human impact of RiskPath is profound. By shifting the perception of healthcare from a reactive service to a proactive one, the toolkit stands to alter how society approaches health management. With a focus on prevention, not only could healthcare costs be curtailed, but patient quality of life could greatly improve through early interventions. As healthcare systems grapple with rising costs and increasing patient loads, the deployment of technologies like RiskPath may serve as a lifeline for making healthcare both effective and efficient.

In summary, the unveiling of RiskPath represents a paradigm shift for predictive healthcare. The combination of advanced artificial intelligence with a commitment to explainable outcomes ensures that patients are not just numbers in a database but individuals whose health journeys can be proactively managed. As the research continues to evolve, one can only imagine the transformative effects this technology could have on every corner of the healthcare landscape.

The full study detailing RiskPath was recently published in the journal Patterns, underscoring the academic rigor and potential real-world applications of this innovative software. With the backing of respected entities such as the National Institute of Mental Health, the research group’s dedication to responsible AI practices reflects a deep commitment to not only advancing technology but doing so in a manner that prioritizes human health and ethical considerations at every step of the way.

Subject of Research: Not applicable
Article Title: RiskPath: Explainable deep learning for multistep biomedical prediction in longitudinal data
News Publication Date: 28-Apr-2025
Web References: RiskPath Study
References: National Institute of Mental Health (grant number R00MH118359)
Image Credits: Kristan Jacobsen Photography / University of Utah Health

Keywords

1. Medical diagnosis
2. Artificial intelligence
3. Risk assessment
4. Decision making

A groundbreaking study spearheaded by seismologists from the University of Utah and the University of New Mexico has harnessed innovative seismic imaging techniques to render unprecedented views of this hidden world. By deploying an extensive network of 650 portable seismometers, known as geophones, spaced 100 to 150 meters apart across Yellowstone’s caldera, and employing a controlled mechanical vibration source, the scientists created high-resolution 2D seismic reflection images. This approach allowed them to pinpoint with remarkable precision the top of the magma chamber at approximately 3.8 kilometers beneath the surface.

The research, recently published in the prestigious journal Nature, marks a major leap forward in volcanic science. Prior to this, most seismological studies relied on naturally occurring earthquakes to map subterranean features, which, while insightful, often yielded blurry and indistinct images—akin to old CT scans. By generating their own seismic “earthquakes” using a Vibroseis truck, which emits controlled ground vibrations usually reserved for oil and gas exploration, the team achieved clarity and detail far surpassing previous efforts.

One of the most critical findings lies in the chamber’s well-defined upper boundary. The magma chamber’s roof is sharply demarcated from the surrounding solid rock strata, a revelation with profound implications for understanding pressure dynamics and gas escape mechanisms. Approximately 86% of the upper portion is solid crystalline rock, while the remaining 14% consists of pore spaces filled half with molten material and half with volatile gases and liquids. This delineation offers valuable insight into how gases such as CO2 and H2O behave within the magma body and the potential explosivity of future eruptions.

Coauthor Jamie Farrell, chief seismologist for the Yellowstone Volcano Observatory, points to the significance of this precise depth measurement. “At 3.8 kilometers, the pressures and conditions dictate how volatile gases exsolve—that is, come out of solution—from magma,” she explains. “If these gases become trapped at depth, they expand rapidly during decompression, often with explosive consequences. Knowing exactly where this boundary lies helps us model those processes with far greater confidence.”

Fortunately, the new data suggests a less alarming conclusion regarding Yellowstone’s immediate volcanic threat. Much of the gas present in the magma escapes gradually through surface geothermal features such as Mud Volcano, preventing dangerous accumulation beneath the surface. Fan-Chi Lin, a geophysics professor affiliated with the University of Utah, elaborates: “These volatiles tend to rise buoyantly and accumulate at the chamber’s top, but if escape pathways exist, they vent safely to the surface, reducing eruption risks.”

Yellowstone’s magma chamber primarily consists of rhyolite, a high-silica igneous rock known for its viscous characteristics and explosive potential when gas is trapped. Spanning roughly 55 by 30 miles laterally and extending down to around 10 miles deep, this body sits atop a deeper, more extensive reservoir of low-silica basalt containing significantly less molten rock—highlighting the complex magmatic stratification beneath the caldera.

The echoes of Yellowstone’s violent past loom large in public consciousness. The volcano’s last cataclysmic eruption roughly 630,000 years ago reshaped the region dramatically, fueling speculation about future blasts. Though the stakes are high, Farrell and colleagues emphasize that the new findings provide reassurance: the volcano shows no signs of imminent eruption, with the sharp delineation and measured volatile content indicative of a system currently in equilibrium rather than buildup.

Key to unlocking these conclusions was the novel method of seismic data acquisition. The team’s use of an artificial vibration source to generate controlled seismic waves transformed the scale and resolution of the collected data. Deploying 650 portable geophones arrayed methodically across the caldera permitted a dense grid of measurements. Over 110 ground vibration points, producing around 20 vibration “treatments” lasting 40 seconds each, generated comprehensive wave data.

Seismic waves travel in two principal types: Primary waves (P-waves) and Secondary waves (S-waves), each interacting uniquely with subsurface materials. The contrasting velocities and attenuations of these waves upon encountering molten rock versus solid matrix allow researchers to discriminate between solid and liquid phases and gauge pore fluid contents. By meticulously analyzing these seismic signatures, the team quantified pore spaces and volatile contents with unprecedented fidelity.

Mike Poland, scientist in charge of the Yellowstone Volcano Observatory, contextualizes the broader significance beyond Yellowstone itself. “This work refines our understanding of the heat engine powering Yellowstone and melt distribution, factors integral for volcanic hazard assessments,” he states. Moreover, he underscores Yellowstone’s role as a geological laboratory, where lessons learned can inform hazard models for other challenging volcanic systems around the globe, such as the Campi Flegrei caldera in Italy and the submerged volcano of Santorini in Greece.

Advances in seismic imaging technology and techniques used in this study mirror the leaps seen in digital photography that sharpen blurred images into crisp snapshots. As Poland notes, “By combining natural earthquake data with new high-resolution active source seismic data, we now have a window into volcanic interiors that was previously unimaginable.” This progress ushers in a new era of volcanic surveillance, offering a powerful toolset for scientists tasked with safeguarding populations living in the shadow of these restless giants.

In conclusion, this landmark study offers an enriched, sharper view into Yellowstone’s magmatic underworld, illuminating the volatile-rich cap that governs gas escape dynamics and eruption potential. It beautifully exemplifies how leveraging technology from energy exploration and deploying massive portable seismic arrays can revolutionize geoscientific investigations. While the awe-inspiring power beneath Yellowstone remains, for now, it is caged by scientific insight, expanding our ability to anticipate and mitigate future volcanic hazards.

Subject of Research: Not applicable
Article Title: A sharp volatile-rich cap to the Yellowstone magmatic system
News Publication Date: 16-Apr-2025
Web References:

• Nature Article DOI
• University of Utah News on Deeper Yellowstone Magma
• Yellowstone Volcano Observatory
  References:
  Dan, C., Song, W., Schmandt, B., et al. “A sharp volatile-rich cap to the Yellowstone magmatic system.” Nature, 16 April 2025. DOI: 10.1038/s41586-025-08775-9
  Image Credits: Jamie Farrell, University of Utah
  Keywords: Yellowstone magma chamber, seismic imaging, volcanic hazard, supervolcano, geophones, Vibroseis, magma volatiles, rhyolite, seismic waves, P-waves, S-waves, Yellowstone Volcano Observatory