(Press-News.org) OAK BROOK, Ill. – In a new special report, researchers address the cybersecurity challenges of large language models (LLMs) and the importance of implementing security measures to prevent LLMs from being used maliciously in the health care system. The special report was published today in Radiology: Artificial Intelligence, a journal of the Radiological Society of North America (RSNA).
LLMs, such as OpenAI’s GPT-4 and Google’s Gemini, are a type of artificial intelligence (AI) that can understand and generate human language. LLMs have rapidly emerged as powerful tools across various health care domains, revolutionizing both research and clinical practice. These models are being employed for diverse tasks such as clinical decision support, patient data analysis, drug discovery and enhancing communication between health care providers and patients by simplifying medical jargon. An increasing number of health care providers are exploring ways to integrate advanced language models into their daily workflows.
“While integration of LLMs in health care is still in its early stages, their use is expected to expand rapidly,” said lead author Tugba Akinci D’Antonoli, M.D., neuroradiology fellow in the Department of Diagnostic and Interventional Neuroradiology, University Hospital Basell, Switzerland. “This is a topic that is becoming increasingly relevant and makes it crucial to start understanding the potential vulnerabilities now.”
LLM integration into medical practice offers significant opportunities to improve patient care, but these opportunities are not without risk. LLMs are susceptible to security threats and can be exploited by malicious actors to extract sensitive patient data, manipulate information or alter outcomes using techniques such as data poisoning or inference attacks.
AI-inherent vulnerabilities and threats can range from adding intentionally wrong or malicious information into the AI model’s training data to bypassing a model’s internal security protocol designed to prevent restricted output, resulting in harmful or unethical responses.
Non-AI-inherent vulnerabilities extend beyond the model and typically involve the ecosystem in which LLMs are deployed. Attacks can lead to severe data breaches, data manipulation or loss and service disruptions. In radiology, an attacker could manipulate image analysis results, access sensitive patient data or even install arbitrary software.
The authors caution that cybersecurity risks associated with LLMs must be carefully assessed before their deployment in health care, particularly in radiology, and radiologists should enact protective measures when dealing with LLMs.
“Radiologists can take several measures to protect themselves from cyberattacks,” Dr. D’Antonoli said. “There are of course well-known strategies, like using strong passwords, enabling multi-factor authentication, and making sure all software is kept up to date with security patches. But because we are dealing with sensitive patient data, the stakes (as well as security requirements) are higher in health care.”
To safely integrate LLMs into healthcare, institutions must ensure secure deployment environments, strong encryption and continuous monitoring of model interactions. By implementing robust security measures and adhering to best practices during the development, training and deployment stages, stakeholders can help minimize risk and protect patient privacy.
Dr. D’Antonoli notes that it is also important to use only the tools that have been vetted and approved by an institution’s IT department, and any sensitive information used as input for these tools should be anonymized.
“Moreover, ongoing training about cybersecurity is important,” she said. “Just like we undergo regular radiation protection training in radiology, hospitals should implement routine cybersecurity training to keep everyone informed and prepared.”
According to Dr. D’Antonoli, patients should be aware of the risks but not overly worried.
“The landscape is changing, and the potential for vulnerability might grow when LLMs are integrated into hospital systems,” she said. “That said, we are not standing still. There is increasing awareness, stronger regulations and active investment in cybersecurity infrastructure. So, while patients should stay informed, they can also be reassured that these risks are being taken seriously, and steps are being taken to protect their data.”
###
“Cybersecurity Threats and Mitigation Strategies for Large Language Models in Healthcare.” Collaborating with Dr. D’Antonoli were Ali S. Tejani, M.D., Bardia Khosravi, M.D., M.P.H., Christian Bluethgen, M.D., M.Sc., Felix Busch, M.D., Keno K. Bressem, M.D., Lisa Adams, M.D., Ph.D., Mana Moassefi, M.D., Shahriar Faghani, M.D., and Judy Wawira Gichoya, M.B.Ch.B., M.S.
Radiology: Artificial Intelligence is edited by Charles E. Kahn Jr., M.D., M.S., Perelman School of Medicine at the University of Pennsylvania, and owned and published by the Radiological Society of North America, Inc. (https://pubs.rsna.org/journal/ai)
RSNA is an association of radiologists, radiation oncologists, medical physicists and related scientists promoting excellence in patient care and health care delivery through education, research and technologic innovation. The Society is based in Oak Brook, Illinois. (RSNA.org)
END
Newly discovered evidence of Australia’s earliest species of tree frog challenges what we know about when Australian and South American frogs parted ways on the evolutionary tree.
Previously, scientists believed Australian and South American tree frogs separated from each other about 33 million years ago.
But in a study published today in Journal of Vertebrate Palaeontology, palaeontologists from UNSW Sydney say the new species, Litoria tylerantiqua, is now at about 55 million years old, ...
New Haven, Conn. — The 2025 Gruber Genetics Prize is being awarded to geneticist and molecular biologist Rotem Sorek, Ph.D., of the Weizmann Institute of Science, for his discoveries in the immune system of bacteria. Using a novel approach that combined computational approaches with an experimental system, Sorek and his colleagues conducted wide scale screens of tens of thousands of bacterial genomes, identifying an astounding number of defense systems used against infection by viruses called phages. This led to ...
New Haven, Conn. — The 2025 Gruber Foundation Cosmology Prize recognizes Ryan Cooke and Max Pettini both for their determination of a key value in the composition of the universe moments after it came into existence and for perfecting the method that allowed them to make that measurement.
Cooke and Pettini will equally share the $500,000 award and each will receive a gold laureate pin at a ceremony that will take place later this year. The citation honors them for “bringing the light element abundances and Big Bang Nucleosynthesis (BBN) into the realm of precision cosmology.”
BBN is a theoretical model of the nuclear reactions in the first ...
New Haven, Conn. — The 2025 Gruber Neuroscience Prize will be awarded to Edward F. Chang, MD, of the University of California, San Francisco (UCSF), in recognition of his pioneering research uncovering how the human brain enables speech. His work has not only revealed how the brain encodes the sounds and movements of spoken language, but has also led to the development of the first successful speech neuroprosthesis to restore communication in individuals with paralysis.
Dr. Chang’s research stands out for both its technical sophistication and its profound clinical impact. It combines high-precision ...
From the IU Innovation and Commercialization Office
Most dental offices are independent practices and often don’t have access to a patient’s medical history. For providing dental care, it is critical that dentists have up-to-date medical and medication histories of their patients to reduce risk during procedures and ensure the success of the treatment.
Thankam Thyvalikakath, DMD, MDS, PhD, professor and associate dean of dental informatics and digital health at the IU Indianapolis School of Dentistry and research scientist at the Regenstrief Institute, and Shuning Li, PhD, M.S., M.S., an assistant research scientist at the IU Indianapolis ...
Postpartum maternal mental health and mother-to-infant bonding are well-established as critical factors in a child’s psychosocial development. However, few studies have explored the combined impact of postpartum maternal depression and early bonding experiences on emotional and behavioral difficulties during middle childhood. A new study reveals significant associations between postpartum depression, mother-to-infant bonding, and child difficulties. Notably, secure early bonding was found to partially buffer the long-term effects of postpartum depression on child outcomes.
The development ...
Background and objectives
Escalating antimicrobial resistance is a global threat, emphasizing the need to explore alternative treatment options. Hence, we aimed to explore the in-vitro activity of ceftazidime-avibactam (CAZ-AVI) in clinical isolates of carbapenem-resistant gram-negative bacteria.
Methods
This was an observational, cross-sectional study conducted at the Microbiology Department of Indus Hospital, Karachi, Pakistan, from January 2023 to October 2024. Carbapenem-resistant gram-negative rods isolated from clinical specimens received from the outpatient, emergency, and inpatient departments were included. Consecutive, ...
FMS-like tyrosine kinase 3 (FLT3) mutations represent some of the most pivotal genetic alterations in acute myeloid leukemia (AML), influencing disease biology, risk stratification, and treatment response. This review highlights the structural, functional, and clinical aspects of FLT3 mutations, emphasizing the transformative role of next-generation sequencing (NGS) in mutation detection, measurable residual disease (MRD) monitoring, and precision therapy.
Introduction
FLT3 encodes a receptor tyrosine ...
Scientists at Northwestern University have developed a new approach that directly combats the progression of neurodegenerative diseases like Alzheimer’s disease and amyotrophic lateral sclerosis (ALS).
In these devastating illnesses, proteins misfold and clump together around brain cells, which ultimately leads to cell death. The innovative new treatment effectively traps the proteins before they can aggregate into the toxic structures capable of penetrating neurons. The trapped proteins then harmlessly degrade in the body.
The “clean-up” strategy significantly ...
When biting into a chili pepper, you expect a fiery sensation on your tongue. This spiciness is detected because of capsaicinoid compounds. But for some peppers, despite high levels of capsaicinoids, the heat is mysteriously dull. Now, researchers reporting in ACS’ Journal of Agricultural and Food Chemistry have identified three compounds that lessen peppers’ pungency. These results challenge the reliability of the century-old Scoville scale, which traditionally bases its rating on two capsaicinoids.
“The discovery of natural dietary compounds that reduce pungency presents promising opportunities ...
