Fortunately, we are not quite there yet, but digitalization is advancing rapidly in the healthcare sector as well. AI chatbots and digital symptom checkers are playing an increasingly important role and are more and more serving as the first point of contact for so-called “self-triage”—that is, the initial assessment of the urgency of treatment by the patients themselves.

But while the technical capabilities of these systems are constantly growing, another factor is coming into the focus of research: how humans communicate with the machine. This is an important topic because even the best technology, especially in medical diagnostics, relies on precise information that users do not always provide in full.

Human reluctance limits the potential of AI

This is the central finding of a study now published in the journal Nature Health. The study was led by Professor Wilfried Kunde, holder of the Chair of Psychology III at the University of Würzburg, and Moritz Reis, a research associate in that department. It involved scientists from Charité – Universitätsmedizin Berlin, the University of Cambridge, as well as Helios Klinikum Emil von Behring and Vivantes Klinikum Neukölln in Berlin.

“The 500 study participants were tasked with writing simulated symptom reports for two common conditions – unusual headaches and flu-like symptoms” describes lead author Moritz Reis the study design. They were led to believe that their reports would be read either by an AI chatbot or a human doctor. The goal was to examine the quality of these reports in terms of their suitability for a medical urgency assessment.

Loss of quality is evident in reduced level of detail

The key finding: When participants believed they were communicating with artificial intelligence, the suitability of their descriptions for an initial medical assessment deteriorated measurably compared to interactions with supposed medical professionals. This effect was even observed among participants who were actually experiencing the relevant symptoms at the time of the survey.

This loss of quality is directly reflected in the level of detail in the reports. While descriptions provided to medical professionals averaged 255.6 characters, those provided to chatbots averaged only 228.7 characters.

Even though a difference of 28 characters may sound small, the research team states that this effect is practically relevant and can result in even high-performance AI models ultimately providing incorrect medical advice. After all, these models also fail to make an accurate medical assessment if patients do not provide all essential information. The success of digital initial assessments depends less on computational power than on the patient’s willingness to provide a detailed description.

Psychological Barriers: Concerns About a “One-Size-Fits-All Diagnosis”

But why are people so hesitant when it comes to machines? A key reason is likely what’s known as “uniqueness neglect.” “Many people assume that AI cannot grasp the individual nuances of their personal situation and instead merely matches standardized patterns,” explains Wilfried Kunde.

In addition, skepticism about algorithms’ diagnostic capabilities, as well as privacy concerns, may lead people to provide abbreviated or vague information. Moritz Reis sums up the human component this way: “If we don’t trust a machine to understand our uniqueness, we may unconsciously withhold the information it would need to provide precise assistance.” This psychological filter can have the effect that medically relevant details never even reach the system, thereby lowering the quality of the diagnosis.

Improving the dialogue with the machine

In the research team’s view, the findings clearly show that the technical advancement of AI alone is not sufficient. They therefore see a potential solution in the intelligent design of user interfaces.

To improve the quality of symptom reports, developers should provide concrete examples of high-quality descriptions and program the AI to actively request missing details. Only when users are encouraged to provide detailed information misdiagnoses can be avoided and the burden on the healthcare system could be effectively reduced.

Continuous piperacillin/tazobactam infusion via elastomeric pump offers a safe, cost-effective alternative to inpatient antibiotic therapy in paediatric oncology, with measurable benefits for ward capacity, healthcare costs and quality of life.

Bacterial infections are among the most common and serious complications in children with cancer, frequently requiring prolonged antibiotic therapy. Piperacillin/tazobactam (pip/taz) is the established first-line intravenous treatment, but its standard three-times-daily dosing schedule creates a significant logistical burden — particularly for families living far from tertiary care centres. A six-month pilot at Oulu University Hospital (OYS) explored whether continuous pip/taz infusion via elastomeric pump could safely shift this therapy out of the inpatient setting, with meaningful benefits for patients, families and the healthcare system.

Background and rationale

OYS serves children across the Northern Finland collaboration area, a large and sparsely populated region where many families cannot realistically attend three outpatient infusion visits per day. Prior to the pilot, pip/taz therapy therefore required full hospitalisation for the entire treatment course — often three to five days, but sometimes several weeks. Elastomeric pumps, well-established in adult oncology, offered an alternative: continuous 24-hour infusion requiring only once-daily pump replacement, enabling home-based treatment or a single daily outpatient visit.

Methods

The pilot ran from November 2024 to April 2025. Weight-based dosing was established for children weighing 15–37.5 kg, using 120 ml FOLfusor (Baxter) pumps for lighter patients and 240 ml pumps for those weighing 30 kg and above. Children weighing 40 kg or more received the standard adult pump containing 12/1.5 g of pip/taz. Piperacillin was reconstituted at 173 mg/ml; tazobactam calculations were unnecessary given the fixed 4:0.5 ratio of the infusion powder. All pumps were prepared centrally at the hospital pharmacy under cleanroom conditions, with full batch documentation for every dose.

Results

The pilot demonstrated clear clinical and operational benefits. First, children were discharged from hospital earlier; once home, families reported improved appetite and increased physical activity in their children. Second, ward workload fell substantially — 117 hospital days were saved over the six-month period. Each elastomeric pump was priced at €95, a figure that covers the medication, the device itself, all required supplies and materials, and pharmacy preparation costs including personnel, cleanroom facilities, and microbiological monitoring. Compared with the cost of inpatient care, this translated to total savings of €54,000–73,000 over the pilot period. Third, pump therapy was successfully delivered to children throughout the collaboration area, including the smallest eligible patients, with centralised pharmacy preparation supporting consistent medication safety.

Implications for practice

These results confirm that elastomeric pump-delivered pip/taz, long used in adults, can be extended effectively to the paediatric oncology population. The model reduces pressure on inpatient beds, lowers nursing workload and generates significant cost savings — while simultaneously improving quality of life for children and their families during an already demanding period of treatment. On the basis of the pilot’s findings, pump-based pip/taz therapy has been adopted as standard practice at OYS Paediatric Haematology and Oncology.

Healthcare professionals seeking further information may contact the OYS Pharmacy team at elina.smolander@pohde.fi or tiina.kallio@pohde.fi.

In an expert commentary titled ‘AI can reason like a physician; what comes next? published in Science, Flinders researchers caution that while new AI systems show impressive capabilities, strong results in controlled studies do not mean they are ready for routine use in hospitals or clinics.

The authors say there is an urgent need to understand how emerging AI tools can be safely integrated into everyday clinical practice, with patient outcomes remaining the central focus.

Despite these warnings, the researchers acknowledge that recent advances in AI create genuine opportunities to support doctors, particularly in busy and high-pressure care settings.

The commentary reviews new research showing that advanced reasoning-based AI systems can work through diagnostic scenarios step by step and, in some cases, closely match or even exceed the diagnostic performance of experienced doctors.

Erik Cornelisse, a PhD candidate at Flinders University and co-author of the commentary, says this shift marks a move from simple question answering tools towards algorithms capable of seemingly human-like clinical reasoning on text-based tasks.

However, the Flinders team stresses that real world medical care involves far more than text-based reasoning or test performance.

They say clinical practice depends on physical examination, listening to patients, understanding medical and social context, and taking responsibility for outcomes, elements that current AI systems cannot safely provide on their own.

“Health care decisions are complex, high stakes, and deeply human, and accuracy alone, particularly on just text-based cases, does not make a system safe for patients,” says Mr Cornelisse from the College of Medicine and
Public Health.

Senior author Associate Professor Ash Hopkins, an NHMRC Investigator and leader of Flinders’ Clinical Cancer Epidemiology Lab, says modern healthcare relies on judgement, accountability, and ethical oversight.

“AI systems have demonstrated that they can reason through clinical problems with similar performance to doctors, notably on the same scenarios used to train clinicians themselves. This presents genuine opportunities to support clinicians in the future,” says Associate Professor Hopkins.

“Multiple stakeholders are currently working on the frameworks for AI in terms of legal, professional, or moral responsibility for its decisions, and presently there is a critical need for deliberate and controlled integration into clinical care.”

The commentary highlights known risks linked to poorly evaluated systems, including bias, inequitable care, and unintended patient harm.

“History shows that algorithms can worsen outcomes when deployed without sufficient safeguards and can amplify problems as easily as they solve them, particularly when systems are trained on incomplete or unrepresentative data,” says Mr Cornelisse.

Looking ahead, the Flinders researchers argue that enthusiasm for medical AI must be matched by strong governance and clearer standards for evaluation.

“We do not allow doctors to practise without supervision and evaluation, and AI should be held to comparable standards,” says Mr Cornelisse.

The researchers stress that improvement in real patient outcomes, not exam scores, benchmarks, or demonstrations, must be the true measure of success.

Associate Professor Hopkins says AI holds enormous promise but must be applied responsibly.

“Patients deserve technology that improves care in the real world, not systems that only look impressive in studies,” he says.

“With careful design, strong oversight, and rigorous evaluation, AI could become a powerful tool to deliver safer, fairer, and more effective care across health systems worldwide,” concludes Associate Professor Hopkins.

The paper, ‘AI can reason like a physician; what comes next?’, by Ashley M. Hopkins and Erik Cornelisse is published in SCIENCE.  DOI 10.1126/science.aeg8766 (link live after embargo lifts)

The findings are based on a nationally representative survey of more than 5,500 U.S. adults conducted from October through December 2025.

In the past 30 days, did you use an AI tool or chatbot for health-related information or advice for any of the following reasons?

% Yes, among adults who have used AI tools or chatbots for health-related information or advice in the past 30 days

Note. Categories are for descriptive purposes only and were not shown on the survey.

Among Americans who have used AI for health-related information or advice in the past 30 days, the most frequently cited motivations are wanting answers quickly (71%) and wanting additional information (71%). Nearly seven in 10 (67%) say they were curious about what AI would say, and roughly six in 10 report using AI to do research on their own before (59%) or after (56%) seeing a doctor.

Regardless of the reason, almost half (46%) of Americans who used AI for healthcare information say the AI tool or chatbot made them feel more confident talking with or asking questions of a provider. Others say it helped them identify issues earlier (22%) or avoid unnecessary medical tests or procedures (19%).

“Artificial intelligence is already reshaping how Americans seek health information, make decisions and engage with providers, and health systems must keep pace,” said Tim Lash, President, West Health Policy Center, a nonprofit and nonpartisan organization focused on aging and healthcare affordability. “The risk isn’t that AI is moving too fast — it’s that health systems may move too slowly to guide its use in healthcare responsibly.”

A Smaller Share Turn to AI in Place of a Provider

While self-directed research is the primary driver of AI health use, a smaller but notable share of recent users report turning to AI instead of seeing a healthcare provider, particularly when faced with cost, access or quality barriers. Among recent AI health users, 27% say they didn’t want to pay for a doctor’s visit and 14% say they were unable to pay. One in five (21%) say they didn’t have time to make an appointment, and 16% say they couldn’t access a doctor or provider. Another 21% say they felt dismissed or ignored by a provider in the past, and 18% say they were too embarrassed to talk to a person.

In the past 30 days, did you use an AI tool or chatbot for health-related information or advice for any of the following reasons?

% Yes, among adults who have used AI for health-related information and advice in the past 30 days

Among recent AI health users, 84% still saw a healthcare provider, but 14% report not seeing a provider they otherwise would have seen because of information or advice they received from AI. When projected to the full U.S. adult population, this represents roughly 14 million Americans who did not see a provider after receiving AI-generated health information.

Trust in that AI-generated health information, however, remains divided. Among those who consulted it in the past 30 days, roughly one-third say they trust it (33%), one-third neither trust nor distrust it (33%), and about one-third distrust it (34%). However, only 4% say they strongly trust the accuracy, indicating that many Americans are making healthcare decisions based on AI-generated information without full confidence in its accuracy.

About one in 10 (11%) who report using AI for health information or advice in the past 30 days say that AI recommended healthcare information or advice they believed was unsafe.

“This data indicates that while some Americans may be using artificial intelligence as a substitute for going to the doctor’s office, many see it as a tool to complement their healthcare, helping them understand symptoms they might be feeling and clarify any diagnosis they receive from their doctors,” said Joe Daly, Global Managing Partner at Gallup.

Motivations Vary by Age and Income

While information-seeking is the dominant reason Americans turn to AI for health purposes, use patterns differ by demographics. Younger adults are more likely than older adults to use AI for self-directed research — 69% of adults aged 18 to 29 say they do research before seeing a doctor, compared with 43% of those 65 and older.

Income differences are most visible in barrier-driven motivations. Among adults earning less than $24,000 annually, 32% say they used AI because they could not pay for a doctor’s visit, compared with just 2% among those earning $180,000 or more.

Everyday Health Questions Top the List of AI Use Cases

Americans who used AI for health information or advice in the past 30 days most often report using it to gather information about everyday health concerns, including physical symptoms (58%) and nutrition or exercise (59%). But AI use extends beyond symptom-checking — Americans who used AI in the past 30 days also report using AI to understand medication side effects (46%), interpret medical information (44%), or research a diagnosis or medical condition (38%). Nearly one in four (24%) report using AI to explore mental health or emotional concerns.

Methodology

West Health-Gallup Center on Healthcare, October-December 2025

Results are based on a Gallup Panel study conducted Oct. 27-Dec. 22, 2025, with a sample of 5,660 adults aged 18 and older who are members of the Gallup Panel, a nationally representative, probability-based panel of U.S. adults. Gallup uses random selection methods to recruit Panel members, including random-digit-dial (RDD) phone interviews that cover landlines and cellphones and address-based sampling (ABS) methods. Respondents with internet access completed the questionnaire as a web survey, and those without regular internet access were sent a printed questionnaire to complete and return by mail. The sample for this study was weighted to be demographically representative of the U.S. adult population, using the most recent Current Population Survey figures. For results based on this sample, one can say that the maximum margin of sampling error is ±2.1 percentage points at the 95% confidence level. Margins of error are higher for subsamples. In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error and bias into the findings of public opinion polls.

“When we negatively interpret a situation, it impacts how we feel and respond—especially in people experiencing anxiety and depression,” said senior author Courtney Beard, PhD,  director of the Cognition and Affect Research Education (CARE) Laboratory at McLean Hospital, a member of the Mass General Brigham healthcare system. “By providing a simple, game-like exercise through an app, we have shown that we can help individuals gain insight into their thinking patterns in a more accessible and engaging way, that leads to meaningful improvements.”

Access to evidence-based treatments for anxiety and depression remains a significant challenge for many individuals due to provider shortages, high costs, and stigma surrounding mental health care. Digital tools have the potential to bridge these gaps; however, most available apps are not rigorously studied, resulting in a wide variance in quality and effectiveness. In addition, users often drop off these apps shortly after download. The researchers designed HabitWorks with these limitations in mind, working with an advisory board of individuals with lived experience of anxiety and depression.

In their new study, the investigators enrolled 340 adults across 44 states, who were randomized to use the HabitWorks app for four weeks or to a control condition that involved self-assessment surveys tracking symptoms of depression and anxiety.

Participants using HabitWorks reported significantly greater improvements in interpretation bias, functioning, and overall mental health symptom severity after one month compared to the control group. HabitWorks also achieved excellent retention rates with 77.8% of participants still using the app in week 4 and 84.4% of participants completing the post-intervention assessment.

“One thing that makes our approach unique in digital mental health is its focus on short, five-minute exercises,” said lead author Alexandra Silverman, PhD, a clinical investigator in the CARE Laboratory. “Unlike traditional interventions that mimic long therapy sessions, HabitWorks aligns with how people use their phones in short bursts, creating an approach that fits into daily life.”

HabitWorks is currently not available to the public. Further research is needed to identify which populations would benefit most from HabitWorks, the longevity of its effects and methods for delivering the intervention beyond a research setting. For more information on HabitWorks and to sign up for its waitlist, visit this website.

Authorship: In addition to Silverman and Beard, Mass General Brigham authors include Gabriela Kovarsky Rotta and Doah Shin.
Disclosures: None.
Funding: This work was supported by the National Institute of Mental Health (R01MH12937) and by Harvard Medical School’s Livingston Fellowship and McLean Hospital’s Pope-Hintz Endowed Fellowship.
Paper cited: Silverman, A. et al. “Randomized Controlled Trial of Smartphone-Based Interpretation Bias Intervention for Anxiety and Depression” Journal of Consulting and Clinical Psychology DOI: xxx

The authors show that existing regulatory requirements for approval have so far only partially addressed many of these so-called “human factors-related risks”. This can create gaps that impact the safety and quality of care. To address these, the researchers identify seven key risks and develop practical recommendations for action that can be integrated into existing regulatory and documentation processes.

Risks in the use of AI systems

AI-based medical devices can be used in various areas of clinical environments. In radiology, for example, they assist in detecting cancer. Clinical decision support systems help select personalized therapies for patients.  AI can also support real-time monitoring and early warning systems, as well as chatbots for applications such as patient communication and software that automatically generate medical reports or summarize findings. The analysis focuses on risks that may arise in the practical use of such AI systems. These include, for example, an increased likelihood of outputs being misunderstood or misinterpreted due to the sometimes-opaque nature of AI systems. Problems can also occur when trust in the application is miscalibrated: resulting in users either relying too heavily on AI assistance or ignoring relevant recommendations. The researchers also point to the risk of automation bias: the tendency to uncritically adopt recommendations from automated systems, potentially overlooking errors or forgoing independent judgment. Additional risks include potential deskilling, technostress among users, an unchecked expansion of indications beyond the originally intended scope (indication creep), and errors related to system changes or different operating modes. Such factors can create additional burdens or unexpected failures in clinical practice – even when the technical performance of a system itself is strong.

A practical guide for manufacturers and evaluators

For their analysis, the research team evaluated existing standards on usability and safety, regulatory guidelines, alongside the scientific literature on AI in healthcare. In addition, expert discussions from the fields of clinical application, regulation, and human factors were incorporated. The result is a practical guide, that fills a gap in current standards, with seven recommendations. These are intended to support manufacturers and evaluators both before and after a product is placed on the market. The aim is to identify AI-specific risks in interaction with human users at an early stage and to address them systematically.

The framework recommends developing and deploying AI-based medical devices in a way that clearly defines the users, in which context the systems are applied, and which tasks are assigned to humans and which to the system. Furthermore, results should be presented in a way that is easy to understand, integrated into existing clinical workflows, and supplemented by training where needed as well as safe fallback options in the event of system failures. The authors emphasize the importance of continuous monitoring after market entry. Usage patterns, potential misuse, or overreliance on AI systems should be systematically observed and corrected as needed. Changes to the systems must also be communicated transparently so that work processes can be adjusted accordingly.

The recommendations are deliberately formulated in general but regulatory-aligned terms so that they can be applied to different AI-enabled medical devices and application scenarios. In a next step, the researchers aim to test and further develop their recommendations based on concrete pilot applications with AI-enabled medical devices. In the long term, human factors should be systematically considered in the regulation and evaluation of AI-based health technologies – reducing avoidable risks while supporting safe innovation in medicine.

The article was authored by researchers from TU Dresden (EKFZ for Digital Health, Chair of Industrial Design Engineering, and Faculty of Business and Economics), in collaboration with experts from the University of Oxford (United Kingdom) and Geneva University Hospital (Switzerland).

Publication

Rebecca Mathias, Anne Schmitt, Mateo Campos, Baptiste Vasey, Sebastian Lorenz, Peter McCulloch, Stephen Gilbert: Evaluation of Human Factors-Related Risks in AI-Enabled Medical Devices: A Practical Guide, NEJM AI, 2026. Link: https://ai.nejm.org/doi/full/10.1056/AIpc2501297

Else Kröner Fresenius Center (EKFZ) for Digital Health

The EKFZ for Digital Health at the Faculty of Medicine at TUD Dresden University of Technology and University Hospital Carl Gustav Carus Dresden was established in September 2019. It receives funding of around 40 million euros from the Else Kröner Fresenius Foundation for a period of ten years. The center focuses its research activities on innovative, medical and digital technologies at the direct interface with patients. The aim here is to fully exploit the potential of digitalization in medicine to significantly and sustainably improve healthcare, medical research and clinical practice.

The World Health Organization estimates that 1 in 10 medications ranging from cancer treatment to contraceptives are either fake or otherwise “substandard.” Though this primarily affects the developing world, there are also gray markets for weight-loss or anti-aging drugs in the U.S.

“Watered-down or illicit versions of drugs like Botox or popular GLP-1 inhibitors have caused serious injuries or death,” said William Grover, associate bioengineering professor at the University of California, Riverside.

In response to this problem, Grover’s laboratory has developed a fake drug detector that could be manufactured for under $30, and potentially for as little as $5. Open-source plans to build the device are detailed in a new paper in the journal Analytical Chemistry.

At its core is a low-cost infrared sensor made for use in toy robots able to follow lines drawn on paper. The researchers repurposed the sensors to instead track the rate at which pills dissolve in water.

All pills of a given drug dissolve — or should dissolve — at roughly the same rate.  Legitimate medications don’t necessarily dissolve any faster or slower than counterfeit ones. But they were made by different people at different facilities and with different ingredients, so their dissolution rates form a “fingerprint” that makes them identifiable and different from that of a fake drug.

“The theory here is that if it’s a legitimate medicine, the manufacturer made every pill identical enough that they’ll all behave roughly the same way when they dissolve,” Grover explained. “So if you test a suspect pill, and it dissolves at a different rate than the real thing, this suggests the suspect pill is counterfeit.”

While others have used dissolution rates to determine a medication’s legitimacy, Grover’s laboratory made the tests more sophisticated by creating an electronic device that converts a pill’s dissolution into a digital signature that they call a “disintegration fingerprint.”

After designing the device, the researchers sought to create a library of these fingerprints that could be used to identify a suspect pill. The group tested over 30 different medications ranging from antibiotics and vitamin supplements to prescription opioids and over-the-counter painkillers. They found that 90% of these pills could be correctly identified using the fingerprinting method.

The group also tested whether their technique could distinguish name-brand and generic versions of the same drug.

“We took Bayer aspirin pills and drug-store-brand aspirin — these are basically identical medicines with the same active ingredient and very similar inactive ingredients,” Grover said, “but when ran through our tests, we could easily tell the difference between the two products.”

The research team even recruited their friends and family to collect samples of drug products from across the U.S. and Canada. They found that pills of the same product typically have similar disintegration fingerprints regardless of where they were purchased. However, some manufacturers make slightly different versions of products for different countries and fingerprinting successfully distinguished U.S. and Canadian versions of a product.

Though there are high-quality pharmaceuticals widely available in the U.S., the CDC warns that there is a public health risk for people ordering what they believe to be prescription medications from disreputable online pharmacies. These medications are frequently found to be fakes.

Other times, a medication could contain irregularities because of manufacturing mistakes. “A facility could get a drum of mislabeled ingredients that can get incorporated into the medicine,” Grover said. “But even an honest error can lead to death.”

In the future, Grover would like to use this method to detect fake antimalarial drugs. These are drugs that treat malaria, a major cause of death in many tropical regions.  Malaria is treatable with the right medications.

“Unfortunately, bad actors know they can make money preying on the need for antimalarials. They sell pills that have the same packaging as authentic antimalarials, but  don’t contain the active ingredients,” Grover said. “If someone gives these pills to their child, they won’t cure their infection.”

Grover hopes to get his tool into the hands of those who can use it to fight fake antimalarials and other fake drugs.

“I can’t imagine a more despicable person than someone who would sell fake medicine to a child. I hope our work makes those criminals’ lives a little harder.”

Image: By counting the particles formed when a pill dissolves in a water-filled cup, the team’s device can identify fake medications.

View more Credit: William Grover/UCR

The researchers in the School of Pharmacy have created a customizable wound scaffold that delivers natural, biodegradable antibacterials over time to encourage healing. Researchers Michael Repka, distinguished professor of pharmaceutics and drug delivery; Sateesh Vemula, postdoctoral researcher; and doctoral candidate Nouf Alshammari published their results in the European Journal of Pharmaceutics and Biopharmaceutics.

“People with limited mobility or diabetes often have wounds with reduced oxygen supply,” Vemula said. “This can slow the body’s normal repair process and make wounds more likely to become long-lasting, while also increasing the chance that bacteria can grow and lead to infection.”

Chronic wounds, including diabetic ulcers and pressure sores, can linger for months or even years.

Repka and his team are 3D-printing a breathable, patch-like structure that can be placed over the wound. The patch is made using chitosan – a natural material found in crustaceans, insects and fungi – along with plant-derived antimicrobials that help fight germs. Chitosan helps accelerate the growth of skin cells while reducing inflammation and preventing infection.

This structure acts as a scaffold, encouraging growth while also protecting the wound from outside sources of infection or contamination.

“A lot of bandages are made with organic solvents, which actually hurt the wound-healing process, especially when applied intimately on the wound,” Repka said. “With the materials and technique we’re using, you don’t have organic solvents.

“We’re also not using traditional antibiotics over a long period of time, because that can often cause the bacteria to become resistant. That’s the advantage of using natural products.”

Using a 3D printer to create the scaffold means that the patch can be tailored to fit any wound on any part of the body.

“The materials we used are also biodegradable,” Alshammari said. “With time, the scaffold is going to be absorbed into the skin. And it’s an inactive material, so we don’t have to worry about side effects or toxic residuals.”

Being biodegradable also means that if the material is applied to wounds inside the body, health care professionals don’t have to make a second incision to remove it, Vemula said.

The technology can be applied to other types of wounds where a traditional bandage would not be suitable, the Ole Miss researchers said.

“Depending on what kind of wound it is, a regular bandage might work well and this wouldn’t be necessary,” Repka said. “But there are a lot of applications for this technology. These could be printed in the field for, say, military applications.

“If you have a generator that can run these 3D printers, you can print the scaffold you need based on what kind of wound has occurred.”

Before the scaffold can be used clinically, it will need further testing and review by the Food and Drug Administration.

“The goal is translating this from research to patients,” Repka said.

Image: Michael Repka, distinguished professor of pharmaceutics and drug delivery, works with a 3D-printed medical device in his lab in Shoemaker Hall. Repka’s latest research shows that 3D-printed wound scaffolding could aid in healing chronic wounds. 

View more Credit: Photo by Thomas Graning/Ole Miss Digital Imaging Services

A new study, published in the March 9 online issue of npj Health Systems [https://doi.org/10.1038/s44401-026-00077-0], suggests that the answer depends less on the AI itself and more on how it’s designed.

The investigators found that health care AI systems work far better when tasks are distributed among multiple specialized AI “agents”—software systems that can perform complex tasks, learn, and adapt—rather than relying on a single, all-purpose agent. This multi-agent approach kept performance steady even as demands increased, while dramatically reducing computing costs and delays, say the investigators.

“For health care organizations, our findings point to a smarter way to use AI,” says senior study author Girish N. Nadkarni, MD, MPH, Barbara T. Murphy Chair of the Windreich Department of Artificial Intelligence and Human Health, Director of the Hasso Plattner Institute for Digital Health, Irene and Dr. Arthur M. Fishberg Professor of Medicine at the Icahn School of Medicine, and Chief AI Officer of the Mount Sinai Health System. “By assigning different tasks, such as finding patient information, extracting data, or checking medication doses, to specialized AI agents, systems can run faster and more reliably while keeping costs under control. Ultimately, this kind of design could help health care teams spend less time on administrative work and more time focusing on patients.”

As part of the study, the researchers compared two approaches to clinical AI: a single system responsible for handling many different clinical tasks, and a coordinated network of specialized AI agents overseen by a central “orchestrator.” Using state-of-the-art language models, the team evaluated performance across common clinical functions, including information retrieval, data extraction, and medication dosing calculations—under simulated real-world conditions involving up to 80 simultaneous tasks.

“What we found is that AI systems behave a lot like people,” says study lead author Eyal Klang, MD, formerly with the Icahn School of Medicine. “When you ask one system to do too many different things at once, performance suffers. But when one orchestrator agent divides the work among specialized agents, the system stays accurate, responsive, and far more efficient, even under heavy demand.”

The coordinated multi-agent system maintained superior accuracy levels while using far fewer computing resources, up to 65 times fewer, than a single-agent design. The study simulated real clinical “traffic,” where many types of tasks arrive at once and compete for attention, the investigators say.

“Our findings show that smart coordination is not just a technical preference,” Dr. Klang says. “It can make the difference between an AI system that continues to function smoothly and one that begins to break down when it is exposed to the pressures of real clinical workloads.”

Next, the research team plans to test these coordinated AI systems directly in clinical settings, using real-time patient data. If successful, this approach could help shape how hospitals and health systems scale AI in the future, helping them handle peak workloads without sacrificing quality or safety.

The researchers emphasize that the gains are not automatic: even sophisticated AI can fall short when systems are poorly designed or implemented. “Health care does not operate one task at a time,” Dr. Nadkarni says. “Hospitals face constant, overlapping demands, especially during busy periods. Our findings show that the future of health care AI is not a single super-intelligent system, but a coordinated team of focused agents that work together to scale safely, control costs, and support real clinical operations.”

“When a single agent handles everything, you can’t trace where it went wrong. With the orchestrator, every step is logged, which tool was called, what it returned, and how the answer was assembled. At 80 simultaneous tasks, the single agent dropped to 16 percent accuracy while burning 65 times more compute—and you’d have no way to figure out why. That kind of transparency isn’t optional in medicine,” says second author Mahmud Omar, MD, a visiting researcher in the Windreich Department. “This matters more now than ever—agentic AI is no longer a research concept. Tools like OpenAI’s operator mode, Claude’s Cowork, and similar platforms are putting autonomous agents directly in the hands of clinicians and patients. As that adoption accelerates, the architecture behind these systems has to be auditable from the start.”

The paper is titled “Orchestrated multi agents sustain accuracy under clinical‑scale workloads compared to a single agent.”

The study’s authors, as listed in the journal, are Eyal Klang, Mahmud Omar, Ganesh Raut, Reem Agbareia, Prem Timsina, Robert Freeman, Lisa Stump, Alexander Charney, Benjamin S. Glicksberg, Girish N. Nadkarni.

This work was supported  in part by the Clinical and Translational Science Awards (CTSA) grant UL1TR004419 from the National Center for Advancing Translational Sciences. Research reported in this publication was also supported by the Office of Research Infrastructure of the National Institutes of Health under award number S10OD026880 and S10OD030463.

For more Mount Sinai artificial intelligence news, visit: https://icahn.mssm.edu/about/artificial-intelligence.

In a new study published in Proceedings of the National Academy of Sciences, the team directly measured how water interactions influence nanoparticle biological performance.

“Water is necessary for all life,” said Navrotsky, the lead author of the study, Regents Professor in the School of Molecular Sciences and director of Arizona State University’s Center for Materials of the Universe. “And in medicine, it is the first molecule that interacts with any nanoparticle surface in a biological environment. By directly measuring the energetics of water adsorption, we can quantify the interaction potential of the nanoparticle surface and better predict how it will behave in the body.”

This so-called hydration energetics were measured for a series of biomolecule-coated magnetite nanoparticles, revealing how different surface coatings alter water interactions, immune recognition and drug delivery potential.

The study, led by Navrotsky and ASU scientists including first author Kristina Lilova, Tamilarasan Subramani, Isabella Montini, Anne Harrison, Manuel Scharrer, Jun Wu and Hongwu Xu, provides the first quantitative, thermodynamic framework linking how primary water energetics relate to nanoparticle biological performance.

Why water matters

Despite major efforts, the promise of nanomedicine has largely failed to deliver a new generation of better drugs to treat illness and disease. This has been mainly due to the human body, which has provided a formidable maze of barriers and defenses for scientists to overcome to deliver the right drug to the right target at the right time.

That’s also why cancer chemotherapy has had its long-known severe side effects, delivering unwanted toxins throughout the body while trying to kill the tumor.

Therefore, scientists have been hard at work to develop a Trojan horse type of nanomedicine therapy by surrounding drugs within a protective cage of nanoparticles.

But there are huge, unresolved challenges.

These nanoparticles, designed for drug delivery, imaging and therapeutic applications must first function in complex biological fluids such as blood, gut or brain fluids after being swallowed. Once introduced into the body, nanoparticles are immediately surrounded by water molecules and biomolecules, forming a nanocomplex stew that dictates their stability, circulation time, immune response and cellular uptake.

Despite the central role of hydration in nanomedicine, previous research had not directly measured the energetics of water adsorption on biomolecule-coated magnetic nanoparticles.

Getting to the core of the problem

The ASU team addressed this gap by studying core–shell nanocomplexes composed of magnetite (iron oxide) cores coated with three representative biomolecules: a protein (bovine serum albumin), a polysaccharide (potato starch) and a fatty acid (lauric acid).

Using a highly sensitive calorimetry–gas adsorption system, the researchers measured the energetics of water adsorption on dry coated nanoparticles, their hydrophilic area and interaction potential, and compared the results to free biomolecules and uncoated magnetite.

The results showed that each coating dramatically alters the hydration behavior—and the biological interaction potential of the nanocomplex.

Patchy protein power

The first experiment of used a nanoparticle coated with a protein, bovine serum albumin (BSA), commonly used as a model for human serum albumin in drug delivery research. Overall, the protein coating produced the strongest initial interaction with water when coated onto magnetite nanoparticles. The BSA-coated particles exhibited strong binding sites exposed at the surface.

However, the total water uptake was lower than that of free BSA, revealing incomplete surface coverage and the presence of uncoated magnetite patches.

“The protein coating increases the surface interaction potential of the nanocomplex,” Lilova explained. “But the existence of exposed magnetite regions introduces heterogeneity that may promote protein corona formation and immune recognition.”

But such “patchiness” could favor the adsorption of opsonins—proteins that tag foreign particles for immune clearance—potentially reducing circulation lifetime.

A starch shell

In contrast, the starch-coated magnetite exhibited a large, water-loving (hydrophilic) surface area, but weaker interaction potential compared to free starch.

The researchers found that starch chains bind to the magnetite surface via hydroxyl groups, reducing the number of groups available for water interaction. Transmission electron microscopy revealed a dense encapsulating shell, limiting accessibility to external water molecules.

“The weaker interaction potential of the starch coating and its relatively large hydrophilic surface area suggest more dynamic and reversible binding,” Lilova said. “This may be beneficial in drug delivery, where mobility along cell membranes and reduced cytotoxicity are desirable.”

Such reversible interactions may allow nanoparticles to engage cell membranes without causing significant disruption—an important consideration for biocompatibility.

Fatty flavor

Perhaps the most striking finding involved lauric acid, a fatty acid coating. Free crystalline lauric acid does not adsorb water, because as any cook knows, water and fatty oils do not mix.  However, when coated onto magnetite nanoparticles, the fat coating reorganized into a partial bilayer structure, resulting in strong water interaction and a stable hydrated interfacial layer.

“The fatty acid rearranges into a partial bilayer with very strong hydrophilicity,” said Lilova. “That structure increases stability and may reduce immune activation compared to more hydrophobic surfaces.”

The bilayer arrangement may also promote longer circulation times in the body.

A better framework for nanomedicine

Across all three coatings, the study establishes that the science of water energetics (hydration enthalpy) can be a key thermodynamic parameter that reflects surface hydrophilicity, heterogeneity and biological interaction potential.

The results from the three coating may help the scientists with a “Goldilocks” predictive tool for getting nanoparticle design “just right.”

“Our findings show that surface functionalization doesn’t just change chemistry—it fundamentally alters the thermodynamic landscape at the nano-bio interface,” said Lilova.

“By understanding primary hydration energetics, we can rationally engineer nanocarriers with tailored stability, immune interactions and drug delivery behavior.”

Looking ahead

The work has broad implications for the design of nanomedicines used in applications such as targeted drug delivery, body imaging contrast agents, cancer treatments and biosensing applications.

“This research provides a thermodynamic foundation for designing nanocarriers with predictable biological reactivity,” said Navrotsky. “It moves us one step closer to truly rational nanomedicine.”

As nanomedicine research continues to evolve, hydration energetics may become a central tool in engineering safer, longer circulating and more effective nanoparticle therapies that could one day save lives. The work also provided a stepping stone for future research focused on the direct measurements of the stabilization effect of representative biomolecular coatings on the nanocomplex.

The research was supported by the U.S. Department of Energy and conducted at Arizona State University’s Center for Materials of the Universe, led by Navrotksy.

Using a large language model (LLM), the MIT team analyzed the genetic code of the industrial yeast Komagataella phaffii — specifically, the codons that it uses. There are multiple possible codons, or three-letter DNA sequences, that can be used to encode a particular amino acid, and the patterns of codon usage are different for every organism.

The new MIT model learned those patterns for K. phaffii and then used them to predict which codons would work best for manufacturing a given protein. This allowed the researchers to boost the efficiency of the yeast’s production of six different proteins, including human growth hormone and a monoclonal antibody used to treat cancer.

“Having predictive tools that consistently work well is really important to help shorten the time from having an idea to getting it into production. Taking away uncertainty ultimately saves time and money,” says J. Christopher Love, the Raymond A. and Helen E. St. Laurent Professor of Chemical Engineering at MIT, a member of the Koch Institute for Integrative Cancer Research, and faculty co-director of the MIT Initiative for New Manufacturing (MIT INM).

Love is the senior author of the new study, which appears this week in the Proceedings of the National Academy of Sciences. Former MIT postdoc Harini Narayanan is the paper’s lead author.

Codon optimization

Yeast such as K. phaffii and Saccharomyces cerevisiae (baker’s yeast) are the workhorses of the biopharmaceutical industry, producing billions of dollars of protein drugs and vaccines every year.

To engineer yeast for industrial protein production, researchers take a gene from another organism, such as the insulin gene, and modify it so that the microbe will produce it in large quantities. This requires coming up with an optimal DNA sequence for the yeast cells, integrating it into the yeast’s genome, devising favorable growth conditions for the it, and finally purifying the end product.

For new biologic drugs — large, complex drugs produced by living organisms — this development process might account for 15 to 20 percent of the overall cost of commercializing the drug.

“Today, those steps are all done by very laborious experimental tasks,” Love says. “We have been looking at the question of where could we take some of the concepts that are emerging in machine learning and apply them to make different aspects of the process more reliable and simpler to predict.”

In this study, the researchers wanted to try to optimize the sequence of DNA codons that make up the gene for a protein of interest. There are 20 naturally occurring amino acids, but 64 possible codon sequences, so most of these amino acids can be encoded by more than one codon. Each codon corresponds to a unique transfer RNA (tRNA) molecule, which carries the correct amino acid to the ribosome, where amino acids are strung together into proteins.

Different organisms use each of these codons at different rates, and designers of engineered proteins often optimize the production of their proteins by choosing the codons that occur the most frequently in the host organism. However, this doesn’t necessarily produce the best results. If the same codon is always used to encode arginine, for example, the cell may run low on the tRNA molecules that correspond to that codon.

To take a more nuanced approach, the MIT team deployed a type of large language model known as an encoder-decoder. Instead of analyzing text, the researchers used it to analyze DNA sequences and learn the relationships between codons that are used in specific genes.

Their training data, which came from a publicly available dataset from the National Center for Biotechnology Information, consisted of the amino acid sequences and corresponding DNA sequences for all of the approximately 5,000 proteins naturally produced by K. phaffii.

“The model learns the syntax or the language of how these codons are used,” Love says. “It takes into account how codons are placed next to each other, and also the long-distance relationships between them.”

Once the model was trained, the researchers asked it to optimize the codon sequences of six different proteins, including human growth hormone, human serum albumin, and trastuzumab, a monoclonal antibody used to treat cancer.

They also generated optimized sequences of these proteins using four commercially available codon optimization tools. The researchers inserted each of these sequences into K. phaffii cells and measured how much of the target protein each sequence generated. For five of the six proteins, the sequences from the new MIT model worked the best, and for the sixth, it was the second-best.

“We made sure to cover a variety of different philosophies of doing codon optimization and benchmarked them against our approach,” Narayanan says. “We’ve experimentally compared these approaches and showed that our approach outperforms the others.”

Learning the language of proteins

K. phaffii, formerly known as Pichia pastoris, is used to produce dozens of commercial products, including insulin, hepatitis B vaccines, and a monoclonal antibody used to treat chronic migraines. It is also used in the production of nutrients added to foods, such as hemoglobin.

Researchers in Love’s lab have started using the new model to optimize proteins of interest for K. phaffii, and they have made the code available for other researchers who wish to use it for K. phaffii or other organisms.

The researchers also tested this approach on datasets from different organisms, including humans and cows. Each of the resulting models generated different predictions, suggesting that species-specific models are needed to optimize codons of target proteins.

By looking into the inner workings of the model, the researchers found that it appeared to learn some of the biological principles of how the genome works, including things that the researchers did not teach it. For example, it learned not to include negative repeat elements — DNA sequences that can inhibit the expression of nearby genes. The model also learned to categorize amino acids based on traits such as hydrophobicity and hydrophilicity.

“Not only was it learning this language, but it was also contextualizing it through aspects of biophysical and biochemical features, which gives us additional confidence that it is learning something that’s actually meaningful and not simply an optimization of the task that we gave it,” Love says.

###

The research was funded by the Daniel I.C. Wang Faculty Research Innovation Fund at MIT, the MIT AltHost Research Consortium, the Mazumdar-Shaw International Oncology Fellowship, and the Koch Institute.

The global scoping review looked at this early phase of testing and revealed huge variations in the way the trials are being conducted and the measures used to assess the effectiveness of the tools.

“This lack of guidance around silent trials is concerning as AI models can be unpredictable and difficult to use in real-world settings if they haven’t been tested thoroughly,” said corresponding author Lana Tikhomirov, a PhD candidate from Adelaide University’s Australian Institute for Machine Learning.

“Some of the trials in our review focused on AI metrics that weren’t clinically useful, while others looked at the bare minimum with no details on how the model performed in a clinical setting.

“If these AI tools are rolled out without comprehensive testing and things go wrong, it could expose both patients and clinicians to harmful advice.”

Silent trials are when AI models are tested in their intended setting for use, but the results don’t influence patient care as they aren’t given to the clinical team at the time of treatment.

Currently there are no formal guidelines on how to conduct these trials, which researchers say are critical to ensure an AI tool will be useful and beneficial in a local setting.

“Silent trials are a low-risk way to test technology without compromising patient outcomes,” said co-author Associate Professor Melissa McCradden, who is the Deputy Director of Adelaide University’s Australian Institute for Machine Learning, AI Director at the Women’s and Children’s Health Network and Hospital Research Foundation Fellow in Paediatric AI Ethics.

“We know that many AI models fail when they’re introduced into real-world settings and an AI tool that works in one hospital may not work in another.

“Conducting comprehensive silent trials that adhere to a clear set of international guidelines is critical if we want to successfully take AI tools from bench to bedside.”

The scoping review has been published in Nature Health and is part of a larger study looking at silent phase evaluations for healthcare AI.

Project CANAIRI – Collaboration for Translational Artificial Intelligence Trials – is focusing on developing guidance for silent trials to ensure health settings in which AI is intended to be used are ready and able to do so in a beneficial way.

“Ultimately, we would like to see silent trials become a mandatory part of the process of adopting AI tools in medicine,” said Associate Professor McCradden, who is the lead of Project CANAIRI.