A groundbreaking new medical dataset is poised to revolutionize healthcare in Africa by improving chatbots’ understanding of the continent’s most pressing medical issues and increasing their awareness of accessible treatment options.

AfriMed-QA, developed by researchers from Georgia Tech and Google, could reduce the burden on African healthcare systems. 

The researchers said people in need of medical care file into overcrowded clinics and hospitals and face excruciatingly long waits with no guarantee of admission or quality treatment. There aren’t enough trained healthcare professionals available to meet the demand.

Some healthcare question-answer chatbots have been introduced to treat those in need. However, the researchers said there’s no transparent or standardized way to test or verify their effectiveness and safety.

The dataset will enable technologists and researchers to develop more robust and accessible healthcare chatbots tailored to the unique experiences and challenges of Africa. 

One such new tool is Google’s MedGemma, a large-language model (LLM) designed to process medical text and images. AfriMed-QA was used for training and evaluation purposes.

AfriMed-QA stands as the most extensive dataset that evaluates LLM capabilities across various facets of African healthcare. It contains 15,000 question-answer pairs culled from over 60 medical schools across 16 countries and covering numerous medical specialties, disease conditions, and geographical challenges. 

Tobi Olatunji and Charles Nimo co-developed AfriMed-QA and co-authored a paper about the dataset that will be presented at the Association for Computational Linguistics (ACL) conference next week in Vienna.

Olatunji is a graduate of Georgia Tech’s Online Master of Science in Computer Science (OMSCS) program and holds a Doctor of Medicine from the College of Medicine at the University of Ibadan in Nigeria. Nimo is a Ph.D. student in Tech’s School of Interactive Computing, where he is advised by School of IC professors Michael Best and Irfan Essa.

Focus on Africa

Nimo, Olatunji, and their collaborators created AfriMed-QA as a response to MedQA, a large-scale question-answer dataset that tests the medical proficiency of all major LLMs. That includes Google’s Gemini, OpenAI’s ChatGPT, and Anthropic’s Claude, among others.

However, because MedQA is trained solely on the U.S. Medical License Exams, Nimo said it is not adequate to serve patients in underdeveloped African countries nor the Global South at-large.

“AfriMed-QA has the contextualized and localized understanding of African medical institutions that you don’t get from Med-QA,” Nimo said. “There are specific diseases and local challenges in our dataset that you wouldn't find in any U.S.-based dataset.”

Olatunji said one problem African users may encounter using LLMs trained on MedQA is that they may advise unfeasible treatments or unaffordable prescription drugs.

“You consider the types of drugs, diagnostics, procedures, or therapies that exist in the U.S. that are quite advanced. These treatments are much more accessible, for example in the US, and Europe,” Olatunji said. “But in Africa, they’re too expensive and many times unavailable. They may cost over $100,000, and many people have no health insurance. Why recommend such treatments to someone who can’t obtain them?”

Another problem may be that the LLM doesn’t take a medical condition seriously if it isn’t predominant in the U.S.

“We tested many of these models, for example, on how they would manage sickle-cell disease signs and symptoms, and they focused on other “more likely” causes and did not rank or consider sickle cell high enough as a possible cause,” he said. “They, for example, don’t consider sickle-cell as important as anemia and cancer because sickle-cell is less prevalent in the U.S.”

In addition to sickle-cell disease, Olatunji said some of the healthcare issues facing Africa that can be improved through AfriMed-QA include:

• HIV treatment and prevention
• Poor maternal healthcare
• Widespread malaria cases
• Physician shortage
• Clinician productivity and operational efficiency

Google Partnership

Mercy Asiedu, senior author of the AfriMed-QA paper and research scientist at Google Research, has dedicated her career to improving healthcare in Africa. Her work began as a Ph.D. student at Duke University, where she invented the Callascope, a groundbreaking non-invasive tool for gynecological examinations

With her current focus on democratizing healthcare through artificial intelligence (AI), Asiedu, who is from Ghana, helped create a research consortium to develop the dataset. The consortium consists of Georgia Tech, Google, Intron, Bio-RAMP Research Labs, the University of Cape Coast, the Federation of African Medical Students Association, and Sisonkebiotik.

Sisonkebiotik is an organization of researchers that drives healthcare initiatives to advance data science, machine learning, and AI in Africa.

Olatunji leads the Bio-RAMP Research Lab, a community of healthcare and AI researchers, and he is the founder and CEO of Intron, which develops natural-language processing technologies for African communities.

In May, Google released MedGemma, which uses both the MedQA and Afri-MedQA datasets to form a more globally accessible healthcare chatbot. MedGemma has several versions, including 4-billion and 27-billion parameter models, which support multimodal inputs that combine images and text.

“We are proud the latest medical-focused LLM from Google, MedGemma, leverages AfriMed-QA and improves performance in African contexts,” Asiedu said. 

“We started by asking how we could reduce the burden on Africa’s healthcare systems. If we can get these large-language models to be as good as experts and make them more localized with geo-contextualization, then there’s the potential to task-shift to that.”

The project is supported by the Gates Foundation and PATH, a nonprofit that improves healthcare in developing countries.

Researchers at Georgia Tech have developed a new approach to cancer imaging and therapy.

The study, led by Costas Arvanitis, associate professor in the George W. Woodruff School of Mechanical Engineering and the Wallace H. Coulter Department of Biomedical Engineering, explores a noninvasive method for tracking immune cells in real time using ultrasound. The study was recently published in the journal Nature Communications.

Read the full story on the George W. Woodruff School of Mechanical Engineering website.

As more satellites launch into space, the satellite industry has sounded the alarm about the danger of collisions in low Earth orbit (LEO).  What is less understood is what might happen as more missions head to a more targeted destination: the moon.

According to The Planetary Society, more than 30 missions are slated to launch to the moon between 2024 and 2030, backed by the U.S., China, Japan, India, and various private corporations. That compares to over 40 missions to the moon between 1959 and 1979 and a scant three missions between 1980 and 2000.

A multidisciplinary team at Georgia Tech has found that while collision probabilities in orbits around the moon are very low compared to Earth orbit, spacecraft in lunar orbit will likely need to conduct multiple costly collision avoidance maneuvers each year. The Journal of Spacecraft and Rockets published the Georgia Tech collision-avoidance study in March.

“The number of close approaches in lunar orbit is higher than some might expect, given that there are only tens of satellites, rather than the thousands in low Earth orbit,” says paper co-author Mariel Borowitz, associate professor in the Sam Nunn School of International Affairs in the Ivan Allen College of Liberal Arts.

Borowitz and other researchers attribute these risky approaches in part to spacecraft often choosing a limited number of favorable orbits and the difficulty of monitoring the exact location of spacecraft that are more than 200,000 miles away.

“There is significant uncertainty about the exact location of objects around the moon. This, combined with the high cost associated with lunar missions, means that operators often undertake maneuvers even when the probability is very low — up to one in 10 million,” Borowitz explains. 

The Georgia Tech research is the first published study showing short- and long-term collision risks in cislunar orbits. Using a series of Monte Carlo simulations, the researchers modeled the probability of various outcomes in a process that cannot be easily predicted because of random variables. 

“Our analysis suggests that satellite operators must perform up to four maneuvers annually for each satellite for a fleet of 50 satellites in low lunar orbit (LLO),” said one of the study’s authors, Brian Gunter, associate professor in the Daniel Guggenheim School of Aerospace Engineering. 

He noted that with only 10 satellites in LLO, a satellite might still need a yearly maneuver. This is supported by what current cislunar operators have reported. 

Favored Orbits

Most close encounters are expected to occur near the moon’s equator, an intersection point between the orbit planes of commonly used “frozen” and low lunar orbits, which are preferred by many operators. Other possible regions of congestion can occur at the Lagrangian points, or regions where the gravitational forces of Earth and the moon balance out. Stable orbits in these regions have names such as Halo and Lyapunov orbits. 

“Lagrangian points are an interesting place to put a satellite because it can maintain its orbit for long periods with very little maneuvering and thrusting. Frozen orbits, too. Anywhere outside these special areas, you have to spend a lot of fuel to maintain an orbit,” he said.

Gunter and other researchers worry that if operators aren’t coordinated about how they plan lunar missions, opportunities for collision will increase in these popular orbits.

“The close approaches were much more common than I would have intuitively anticipated,” says lead study author Stef Crum.

The 2024 graduate of Georgia Tech’s aerospace engineering doctoral program notes that, considering the small number of satellites in lunar orbit, the need for multiple maneuvers was “really surprising.”

Crum, who is also co-founder of Reditus Space, a startup he founded in 2024 to provide reusable orbital re-entry services, adds that the cislunar environment is so challenging because “it’s incredibly vast.”

His research also examines ways to improve object monitoring in cislunar space. Maintaining continuous custody of these objects is difficult because a target’s position must be monitored over the entire duration of its trajectory. 

“That wasn’t feasible for translunar orbits, given the vast volume of cislunar orbit, which stretches multiple millions of kilometers in three dimensions,” he says.

By estimating a satellite’s orbit using observed data and constraining the presumed location and direction of the satellite, rather than continuous tracking (a process known as continuous custody), Crum greatly simplified the process. 

“You no longer need thousands of satellites or a set of enormous satellites to cover all potential trajectories,” he explains. “Instead, one or a few satellites are required, and operators can lose custody for a time as long as the connection is reacquired later.”

Since the team started their study, there has been a lot of interest in the moon and cislunar activity — both NASA and China’s National Space Administration are planning to send humans to the moon. In the last two years, India, Japan, the U.S., China, Russia, and four private companies have attempted missions to the moon. 

Why the Moon

Spacefaring nations’ intense interest in exploring the lunar surface comes as no surprise given that the moon offers a variety of resources, including solar power, water, oxygen, and metals like iron, titanium, and uranium. It also contains Helium-3, a potential fuel for nuclear fusion, and rare earth metals vital for modern technology. With the recent discovery of water ice, it could be a plentiful source for rocket fuel that can be created from liquifying oxygen and hydrogen needed to launch deep space missions to destinations like Mars. In February, Georgia Tech announced that researchers have developed new algorithms to help Intuitive Machines’ lunar lander find water ice on the moon.

Commercial space companies like Axiom Space and Redwire Space, as well as space agencies, are actively building lunar infrastructure, from satellite constellations to orbital platforms to support communication, navigation, scientific research, and eventually space tourism. 

A key project involves the Lunar Gateway, a joint venture of NASA and international space agencies like ESA, JAXA, and CSA, as well as commercial partners. Humanity’s first space station around the moon will serve as a central hub for human exploration of the moon and is considered a stepping stone for future deep space missions.

Getting Ahead of a Gold Rush to the Moon

All this activity underscores the urgency to get out in front of potential crowding issues — something that hasn’t occurred in LEO, where near-miss collisions, or conjunctions, are frequent. LEO, which is 100 to 1,200 miles above the Earth’s surface, is host to more than 14,000  satellites and 120 million pieces of debris from launches, collisions, and wear and tear, reports Reuters.

“Using the near-Earth environment as an example, the space object population has gone from approximately 6,000 active satellites in the early 2020s to an anticipated 60,000 satellites in the coming decade if the projected number of large satellite constellations currently in the works gets deployed. That poses many challenges in terms of how we can manage that sustainably,” observed Gunter. “If something similar happens in the lunar environment, say if Artemis (NASA’s program to establish the first long-term presence on the moon) is successful and a lunar base is established, and there is discovery of volatiles or water deposits, it could initiate a kind of gold rush effect that might accelerate the number of actors in cislunar space.”

For this reason, Borowitz argues for the need to begin working on coordination, either in the planning of the orbits for future missions or by sharing information about the location of objects operating in lunar orbit. She pointed out that spacecraft outfitted for moon missions are expensive, making a collision highly costly. Also, debris from such a scenario would spread in an unpredictable way, which could be problematic for other objects.

Gunter agreed, noting, “If we’re not careful, we could be putting a lot of things in this same path. We must ensure we build out the cislunar orbital environment in a smart way, where we’re not intentionally putting spacecraft in the same orbital spaces. If we do that, everyone should be able to get what they want and not be in each other’s way.”

Borowitz says some coordination efforts are underway with the UN Committee on the Peaceful Uses of Outer Space and the creation of an action team on lunar activities; however, international diplomacy is a time-consuming process, and it can be a challenge to keep pace with advancements in technology.

She contends that the Georgia Tech study could provide baseline data that “could be helpful for international coordination efforts, helping to ensure that countries better understand potential future risks.”

Gunter and Borowitz say that follow-on research for the team could involve looking into the Lunar Gateway orbit and other special orbits to see how crowded that space will likely get, and then do an end-to-end simulation of these orbits to determine the most effective way to build them out to avoid collision risks. Ultimately, they intend to develop guidelines to help ensure that future space actors headed to the moon can operate safely.

Candace Washington never thought she’d one day run her own business or teach the next generation of project management leaders in construction and engineering. But that’s exactly what she’s doing thanks to Georgia Tech. 

In 2012, Washington, a seasoned construction veteran with 25 years of expertise and a master’s degree in building construction from Georgia Tech, noticed a shortage of project managers. She oversaw capital improvements and construction buildouts nationally and was consistently getting asked by clients to oversee the construction buildouts. This would spark the idea to start her business and launch Cancave Management & Engineering. 

Over the next decade, Washington built a successful company and yet she continued to see this recurring shortage of project managers. According to Associated Builders and Contractors, the construction sector still grapples with a significant talent shortage that extends beyond the skilled trades to include construction management positions, with a projected need for nearly half a million additional workers in 2025 alone.

“We have fewer people entering the industry. With the pandemic, we had a great exodus where a lot of people decided to get out of the industry and retire early, and then you have the emerging housing market and infrastructure needs, creating demand for construction in general — the perfect storm,” Washington said.

Determined to find more ways to address the problem, she joined Georgia Tech’s School of Building Construction as a part-time instructor and, in 2024, began pursuing her Ph.D. at Tech, where she learned about the Jim Pope Fellowship.

“Being a Pope Fellow has been transformational to my experience as an entrepreneur,” Washington said. “When I started my company, I wish I had something like this. Through this fellowship, I was able to dig deeper into my idea, validate assumptions, and shape it into a solution that addresses the pain points of labor shortages and compliance bottlenecks in the underutilization or over-utilization of resources.” 

As a fellow, Washington was also awarded $15,000 in discretionary funds to support her teaching and entrepreneurial efforts. With the resources from Jim Pope, Washington has been able to make meaningful impacts for students and her company. 

Over the last year, she has worked on the next evolution of her business by building Extend the Ladder®,  a workforce resource and compliance platform built around an industrywide shared resource model for construction professionals. One application of her platform would allow general contractors to share resources by enabling them to find and coordinate talent from a single database.

In addition to helping her pursue a construction job-matching platform, the fellowship has reinforced her love of teaching and mentoring entrepreneurial-minded students. As a part of the fellowship, Washington taught CREATE-X’s Startup Lab, which teaches the fundamentals of evidence-based entrepreneurship.

One student, Vivianne Akerman, a rising junior in industrial engineering, became Washington’s mentee after her spring Startup Lab class. Bitten by the entrepreneurial bug, Akerman decided to continue her entrepreneurial journey in CREATE-X’s Idea-to-Prototype (I2P) course. She turned an idea into action with guidance from Washington, building a solution for a problem she identified during Startup Lab.

“Candace is an amazing mentor who pushes students to be their best selves,” said Akerman, who is developing a makeup platform designed “to make makeup practical and less overwhelming.” The platform will enable consumers to compare and review products and ultimately find what brands work best for them, given their skin type and desired look.

“I love how positive she is,” adds Akerman. “This is new for me — it’s very exciting but also very overwhelming. She helps me stay focused on my priorities and what’s most important.”

Washington emphasizes that there is no guidebook to becoming an entrepreneur; rather, the path must be discovered through conversations, relationship-building, and learning from the experiences of others.

“This experience deepened my appreciation for the spirit of entrepreneurship — it’s been invaluable for me,” she says. “I would tell anybody who's trying to start a business, you need to go through this process.”

Now, as a mentor herself, Washington credits her fellowship in CREATE-X for giving her the confidence and framework to help others. And she credits her path as a mentor and teacher of entrepreneurship to the home she’s found at Georgia Tech. 

Drawing from her own experiences, both the challenges and the triumphs, she offers a piece of advice that she believes aspiring entrepreneurs should carry with them. 

“Start now — you don’t need all the answers. Focus on the process, stay committed, and be open to real-world feedback.”

Applications are now open for the 2025 Jim Pope Fellowship until Sept. 2. Interested faculty can learn more at https://create-x.gatech.edu/faculty/jim-pope-fellowship.

Five teams of researchers from Georgia Tech and Emory University were selected to accelerate their journey from lab to market. Projects include improved cancer detection and therapies, a precise surgical tool and better MRI imaging. Teams will receive funding and commercialization support during the year. Read more about each project here.

The Laser Interferometer Gravitational-Wave Observatory (LIGO)’s LIGO-Virgo-KAGRA (LVK) collaboration has detected an extremely unusual binary black hole merger — a phenomenon that occurs when two black holes are pulled into each other's orbit and combine. Announced yesterday in a California Institute of Technology press release, the binary black hole merger, GW231123, is the largest ever detected with gravitational waves.

Before merging, both black holes were spinning exceptionally fast, and their masses fell into a range that should be very rare — or impossible. 

“Most models don't predict black holes this big can be made by supernovas, and our data indicates that they were spinning at a rate close to the limit of what’s theoretically possible,” says Margaret Millhouse, a research scientist in the School of Physics who played a key role in the research. “Where could they have come from? It raises interesting questions.”

A binary black hole merger absorbs characteristics from both of the contributors, she adds. “As a result, this is not only the most massive binary black hole ever seen but also the fastest-spinning binary black hole confidently detected with gravitational waves.”

“GW231123 is a record-breaking event,” says School of Physics Professor Laura Cadonati, who has been a member of the LIGO Scientific Collaboration since 2002. “LIGO has been observing the cosmos for 10 years now. This discovery underscores that there is still so much that this instrument can help us learn.”

A Cosmic View

The findings challenge current theories on how smaller black holes form, says School of Physics Assistant Professor and LIGO collaborator Surabhi Sachdev. Smaller black holes are the result of supernovae: dying and collapsing stars. During that collapse, explosions can tear apart or eject part of the star’s mass — limiting the size of the black hole that forms.

“Black holes from supernovae can weigh up to about 60 times the mass of our Sun,” she says. “The black holes in this merger were likely the mass of hundreds of suns.”

Because of its size, GW231123 also allowed the team to study the merger in unprecedented detail. “LIGO has observed scores of black hole mergers,” says Cadonati. “Of these, GW231123 has provided us with the clearest view of the ‘grand finale’ of a merger thus far. This adds a new clue to solve the puzzle that are black holes, including their origins and properties.”

“While we saw that our expectations matched the data, the extreme nature of this event pushed our models to their limits,” Millhouse adds. “A massive, highly spinning system like this will be of interest to researchers who study how binary black holes form.”

Decoding a Split-Second Signal

Millhouse and School of Physics Postdoctoral Fellow Prathamesh Joshi used Einstein’s equations for general relativity to confirm LIGO’s detections.

To find black holes, LIGO measures distortions in spacetime — ripples that are created when two black holes collide. These patterns in gravitational waves can be used to find the signature signal of black hole collisions. 

“In this case, the signal lasted for just one-tenth of a second, but it was very clear,” says Joshi. "Previously, we designed a special study to detect these interesting signals, which accounted for all the unusual properties of such massive systems — and it paid off!”

“To ensure it wasn’t noise, the Georgia Tech team first reconstructed the signal in a model-agnostic way,” Millhouse adds. “We then compared those reconstructions to a model that uses Einstein's equations of general relativity, and both reconstructions looked very similar, which helped confirm that this highly unusual phenomenon was a genuine detection.”

Sachdev says that seeing the signal at both LIGO Observatories — placed in Hanford, Washington and Livingston, Louisiana — was also critical. “These short signals are very hard to detect, and this signal is so unlike any of the other binary black holes that we've seen before,” she says. “Without both detectors, we would have missed it.”

A Decade of Discovery

While the team has yet to determine how the original black holes formed, one theory is that they may have resulted from mergers themselves. “This could have been a chain of mergers,” Sachdev explains. “This tells us that they could have existed in a very dense environment like a nuclear star cluster or an active galactic nucleus.” Their spins provide another clue as spinning is a characteristic usually seen in black holes resulting from a merge.

The team adds that GW231123 could provide clues on how larger black holes are formed — including the mysterious supermassive black holes at the center of galaxies.

“Gravitational wave science is almost a decade old, and we're still making fundamental discoveries,” says Millhouse. “It’s exciting that LIGO is continuing to detect new phenomena,  and this is at the edge of what we've seen thus far. There's still so much we can learn.”

The team expects to update their catalogue of black holes in August 2025, which will provide another window into how this exceptionally heavy black hole might fit into the universe, and what we can continue to learn from it.

 

Funding: The LIGO Laboratory is supported by the U.S. National Science Foundation and operated jointly by Caltech and MIT.

Georgia Tech students played a pivotal role in the award-winning Coffee County Memory Project, an oral history initiative that preserves the stories of school desegregation in rural Georgia.

Launched in 2016, the project was supported by the Institute’s Sustainable Communities Summer Internship Program, run by the Center for Serve-Learn-Sustain (now the Center for Sustainable Communities Research and Education), in which students work full time with community partners across Atlanta and Georgia.

Beginning in 2017, trusted advisers contributed to the success of this work, including Vernon E. Jordan Jr., Christopher Lawton, Ann McCleary and G. Wayne Clough. Clough, who served as Georgia Tech’s president from 1994 to 2008, long advocated for public service, community-engaged research, and interdisciplinary teaching and learning.

In 2019, Georgia Tech students and participating interns Brice Minix and Nabil Patel combed through decades of local newspapers, digitized school board records, and conducted interviews with community members who lived in Coffee County during desegregation. In 2020, Kara Vaughan Adams and Bennett Bush transcribed countless interviews. Samina Patel’s contributions in 2020 and 2021 included graphic and web design.

All their work laid the foundation for two virtual museum exhibits: emergingVOICES of Coffee County and Overcoming Segregation: A Journey Through Coffee County’s Forgotten Stories. The latter received the 2023 Award of Excellence from the American Association of State and Local History. Further recognition came this year when the project earned the 2025 Georgia Association of Museums’ Special Project Award for the PLAYBACK & FASTFORWARD seminar series.

T. Cat Ford, Project Director said, “The Serve-Learn-Sustain interns we partnered with from Georgia Tech were all graduates of Coffee High School. Their efforts turbo-charged our work—not only because they worked tirelessly but also because, as they preserved their own history, they offered valuable insights into their lived experience of this legacy.

Click here to learn more about SCoRE’s Sustainable Communities Internship Program.