News and #Research from Amanda Randles's Biomedical Engineering Lab @DukeU. We are using #HPC to address #biomedical questions. #AcademicTwitter #innovation
Come find the Randles Lab at #SC26! We’re sharing how supercomputing helps us simulate billions of red blood cells and reconstruct millions of heartbeats for patient-specific cardiovascular digital twins.
On Tuesday, catch Wendy Wu on the exascale biomedical digital twins panel and meet her at the poster reception to hear about our billion-scale red blood cell simulations.
On Wednesday, I’ll join the medical digital twins panel and give a talk on vascular digital twins at the NVIDIA booth.
See the flyer for times and locations. We’d love to meet you!
New work from Nusrat Sadia Khan! Can fewer simulations capture coronary blood flow as heart rate changes? Our ACM AI Summit paper explores learning representative states for efficient coronary digital twins.
dl.acm.org/doi/abs/10.1145/3…
🎥 Four new seminars, now on YouTube! Explore wearable devices, Bayesian biosignal analysis, AI for clinical research, and forecasting patient trajectories with Duke’s computational & digital health community.
Watch: youtube.com/@dukecomphealth#DigitalHealth#AIinHealthcare
📚 CCDHI CompHealth Corner
Can AI help design combinations of dietary fiber and gut bacteria that support gut health? A new Nature Chemical Biology study uses machine learning and lab experiments to identify promising pairings. From CCDHI member Ophelia Venturelli and team.
Read more: nature.com/articles/s41589-0…
📚 CCDHI CompHealth Corner — What are we reading this week?
Can AI help identify combinations of dietary fiber and gut bacteria that support gut health?
A new study in Nature Chemical Biology, co-authored by Duke’s Ophelia Venturelli and colleagues, explores how machine learning can guide the design of microbial communities with beneficial functions.
The researchers combined AI with laboratory experiments in a design–test–learn cycle: predicting promising combinations of fibers and bacteria, testing them, and using the results to guide the next round.
Their approach identified combinations involving the dietary fiber inulin that promoted production of butyrate, a compound important for intestinal health. The team also tested model-designed combinations in microbial communities derived from human stool, where they produced predictable beneficial outputs.
This work points toward more targeted ways to study and shape the gut microbiome, providing a foundation for future research into personalized dietary and microbial interventions.
Read more and access the paper:
nature.com/articles/s41589-0…
One of the most fun parts of research is getting to share it with others!
Here are Jorik Stoop and Jakob Olsen as the lab gets ready to welcome undergraduates from the Duke Biomedical Engineering Society. The team organized the demos and tour for the students, and I’m proud to have a lab so excited to share their work.
Thanks to everyone in the Randles Lab who helped make it happen!
Please join us tomorrow for our Computational & Digital Health Seminar with Ricardo Henao!
Ricardo will discuss in-context learning and how large language models can use just a few examples to support tasks in healthcare and clinical research. Looking forward to the conversation!
🗓️ Wednesday, November 11 | 2–3 PM Eastern
💻 Virtual and open to everyone
🔗 duke.is/CompHealthNov2026Sem…
📚 CCDHI CompHealth Corner — What are we reading this week?
Can AI recognize when its predictions might be uncertain?
A new study in npj Digital Medicine, co-authored by Evan Calabrese, a member of Duke’s Center for Computational and Digital Health Innovation, explores an AI framework that segments meningiomas on brain MRI while estimating uncertainty in its predictions.
The model achieved high segmentation accuracy, and its uncertainty maps highlighted ambiguous tumor boundaries and other areas where predictions warranted closer review. The researchers also translated this uncertainty into ranges around tumor volume measurements, which could help distinguish meaningful growth or treatment response from measurement variability.
The approach maintained high segmentation performance in external validation involving 353 patients.
This work points toward more transparent AI tools for brain tumor monitoring, with uncertainty estimates that could help clinicians identify results needing closer review.
Read more and access the paper:
comphealth.duke.edu/publicat…#CompHealthCorner#AIinHealthcare#MedicalImaging#DukeUniversity
Join Duke’s Center for Computational and Digital Health Innovation for our next Computational & Digital Health Seminar, featuring Ricardo Henao, Associate Professor of Biostatistics & Bioinformatics at Duke University.
In-Context Learning with Applications in Healthcare and Clinical Research
Dr. Henao will explore how large language models can use examples and instructions to address clinical tasks without retraining, from identifying clinical trial endpoints in medical histories to estimating risk and assessing patient eligibility for trials. He will also discuss the importance of rigorous validation and expert oversight.
📅 Wednesday, November 11, 2026
🕑 2:00–3:00 PM Eastern
💻 Virtual and open to everyone
🔗 duke.is/CompHealthNov2026Sem…
Please share with colleagues and students interested in AI, healthcare, and clinical research!
#DigitalHealth#AIinHealthcare#ClinicalResearch#DukeUniversity
📢 Calling Duke undergraduate researchers!
The Randles Lab is recruiting through Pratt Research Fellows!
Interested in computational modeling, artificial intelligence, and cardiovascular health? Join us to explore how computational methods can help us understand blood flow and how it changes over time.
Project: Computational methods for long-term hemodynamics assessment
Students will analyze blood flow simulation data, explore patterns across activity states, and use statistical and computational tools to study cardiovascular function.
We encourage students interested in AI, computational modeling, and data analysis to apply. Experience with Python and statistics is useful.
📅 Application deadline: October 31
🔗 Apply through Pratt Research Fellows:
pratt.duke.edu/academics/und…
Learn more about our research:
randleslab.pratt.duke.edu/re…
Congratulations to Japneet Kaur Mavi on passing her preliminary examination! 🎉
Japneet’s research combines patient-specific blood flow simulations, wearable data, and cardiovascular digital twins to study how blood flow changes over time and how those changes may inform our understanding of cardiovascular disease.
Her thoughtful presentation and strong responses to challenging questions demonstrated the depth of her knowledge and the care she brings to her research. We’re proud of her hard work and excited to see her dissertation research take shape!
Please join us in celebrating this milestone! 👏
Randles Lab Duke University Duke University Pratt School of Engineering Duke BME #PhD#DigitalTwins#CardiovascularResearch
Want to know what Duke is doing in Computational Health? Check it out at BMES!
Members of Duke’s Center for Computational and Digital Health Innovation are sharing their work through 19 presentations at #BMES2026—from spatial omics and wearable health analytics to vascular simulation, biomaterials, and biotechnology education.
Join presenters from the Randles, Dunn, Hickey, and Grill labs, along with our Associate Director of Education, Cameron Kim, October 22–24 in Orlando.
📍 Find presenters, times, and locations in the flyer.
💡 Stop by the Duke BME booth to explore the Center’s interactive display!
Come meet our researchers and see what we’re building together.
#DukeEngineering#ComputationalHealth#DigitalHealth
Heading to BMES 2026? Come learn what we’re working on in the Randles Lab—from blood flow in carotid arteries to flow through porous scaffolds.
Shea Middleton will present “Quantifying Morphology-Driven Hemodynamic Changes in the Carotid Bifurcation Through Simulated Forward Head Posture,” exploring how changes in head posture can influence carotid artery geometry and blood flow.
Shihab Kabir will present “Identifying Minimum Spanning Sets of Granular Material Distributions to Capture Fluid Flow Behavior,” a poster highlighting an approach to identifying representative flow behavior in porous scaffolds while reducing simulation costs.
You can also explore an interactive display from the Duke Center for Computational and Digital Health Innovation (CDDHI) at the Duke BME booth. Stop by to see our work in action and talk with us about computational and digital health research!
#BMES2026#RandlesLab#DukeBME#CDDHI#ComputationalModeling#BiomedicalEngineering
📚 CCDHI CompHealth Corner — What are we reading this week?
Can AI recognize when its predictions might be uncertain?
A new study published in npj Digital Medicine, co-authored by Evan Calabrese, a member of the Center for Computational and Digital Health Innovation, explores how AI can segment meningiomas on brain MRI while estimating confidence in its predictions.
The model achieved high segmentation accuracy, and its uncertainty maps highlighted difficult tumor boundaries and other areas where predictions were less reliable. The researchers also translated this uncertainty into ranges around tumor volume measurements, which could help distinguish meaningful tumor growth or treatment response from measurement variability.
The approach was externally validated using MRIs from 353 Duke patients, where it maintained high segmentation performance.
This work points toward more transparent AI tools for monitoring brain tumors, with uncertainty estimates that could help clinicians identify predictions needing closer review.
Read more and access the paper: comphealth.duke.edu/publicat…
Duke University | Duke University Pratt School of Engineering | Duke BME
🇩🇪 Randles Lab is heading to FTC 2026 in Germany!
What does it take to make a digital twin reproducible at exascale?
Randles Lab’s Ayman Yousef will present “Towards Trustworthy Exascale Computing: Deterministic Algorithms for Reproducible, GPU-Accelerated Digital Twins” at FTC 2026 in Berlin. The work profiles two parallel approaches to determinism in large scale scientific computing.
Reproducibility helps researchers debug, validate, and share their codes across computing environments.
Join Ayman on October 15, 11:30 AM to 12:45 PM, in Computing Session 1, Rooms 2+3. If you’re attending, come talk with us.
Conference details: saiconference.com/FTC#FTC2026#HPC#GPUComputing#DigitalTwins#Reproducibility
The best way to understand what our labs are working on? See it for yourself.
At our first Demo Day for the Center for Computational and Digital Health Innovation, visitors tried interactive demonstrations, explored posters and short presentations, and talked directly with the researchers behind the work. Students also had a chance to learn about research opportunities and connect with labs.
A big thank you to Samreen Mahmud and Nusrat Sadia Khan, who served as the event’s main organizers, and to all the participating labs and attendees who made the day such a success.
Read Samreen’s reflections on Demo Day:
comphealth.duke.edu/demo-day…