‘Remarkable’ medical discovery for sepsis moves to next phase of human trials

Florey researchers, working with hospital intensive care clinicians, have shown that sodium ascorbate – a pH-balanced formulation of vitamin C – is effective in treating sepsis

Peer-Reviewed Publication

UNIVERSITY OF MELBOURNE

 

IMAGE: 

FLOREY SCIENTISTS MADE A SPECIAL FORMULATION OF SODIUM ASCORBATE TO TREAT SEPSIS.

view more 

CREDIT: SUPPLIED BY THE FLOREY

Scientists at The Florey have proven that a formulation they pioneered alleviates deadly sepsis, with the next phase of clinical trials to start rolling out across Australia next month.  

Promising results from an initial clinical trial at the Austin Hospital in Melbourne, published today in Critical Care, show that sodium ascorbate – a pH-balanced formulation of vitamin C – is effective in treating sepsis. 

Lead investigator Associate Professor Yugeesh Lankadeva said sepsis is notoriously difficult to treat and is often deadly. 

“Sepsis accounts for 35–50 per cent of all hospital deaths. It occurs when the immune system fails to fight off an underlying infection, causing life-threatening falls in blood pressure, multiple organ failure, and death,” Associate Professor Lankadeva said. 

“In our trial at Austin Hospital, patients were given sodium ascorbate into their bloodstream, resulting in promising improvements to multiple organs.” 

Associate Professor Lankadeva, who is research lead of The Florey’s Systems Neuroscience Theme, said the next phase of the $4.9 million government-funded research project will be offered in intensive care units in Adelaide, Melbourne, Perth, Brisbane, Alice Springs and Sydney. 

“We’ll recruit 300 adult septic patients to receive either our formulation or a placebo as well as normal hospital care. The results will help us gather data to determine the effectiveness of our formulation,” Associate Professor Lankadeva said.  

Austin Hospital’s Director of Intensive Care Research, Professor Rinaldo Bellomo, said the previous part of the trial in his department involved 30 adult sepsis patients between October 2020 and November 2022. 

While receiving hospital intensive care, half the patients were randomly allocated to receive sodium ascorbate, and the other half a placebo. 

The study found that patients with sepsis who received the sodium ascorbate treatment: 

• produced more urine, a sign of improved kidney function 

• required less of the clinically used drug, noradrenaline, to restore blood pressure 

• showed signs of improved function in multiple organs. 

“Sepsis is the biggest killer in intensive care units in Australia and worldwide,” Professor Bellomo said. “It often develops so quickly that patients are already critically ill by the time they reach us. A treatment that acts quickly, is safe and highly effective would be an absolute game-changer.” 

Senior Florey researcher on the project, Professor Clive May has spent more than 20 years investigating how sepsis causes organ failure, particularly in the brain and kidneys. 

“Through our showing reduced levels of oxygen in tissues in sepsis, we identified sodium ascorbate as a possible treatment. 

“We’ve seen dramatic results in our work with preclinical studies, where an extremely high dose of sodium ascorbate resulted in full recovery within just three hours, with no side effects. It’s heartening to see all those years of painstaking research pay off with a treatment now within reach for patients,” Professor Clive May said. 

Surviving sepsis: A patient’s perspective  

Long-term Florey staff member Brett Purcell is a consumer representative on the MEGASCORES research program, providing the invaluable perspective of a sepsis survivor. 

"I was rushed to hospital by ambulance in 2011 with a high temperature and delirium. I had the early stages of sepsis. My condition slowly deteriorated and after 12 days I was transferred to a bigger hospital. By then I was in septic shock with a severely infected heart. 

“Six months earlier I had a successful aortic valve replacement; unfortunately the valve had become infected. 

“The surgical team repaired the damage in a six-hour operation. But my condition worsened to critical. My family was told I only had an hour to live as my blood oxygenation levels were not conducive to life. What saved me was the excellent decision making by the surgical team and the ICU intensivists. I was placed on life support on an ECMO machine and given dialysis and my condition rapidly improved. 

“I returned home after almost eight weeks in hospital. 

“I’m so lucky to be alive, and I’m hoping that this new research with sodium ascorbate will provide hospitals with a new life-saving tool, one that’s less invasive, quick and highly effective in fighting sepsis." 

Scientists at The Florey have proven that a formulation they pioneered alleviates deadly sepsis, with the next phase of clinical trials to start rolling out across Australia next month.  

Promising results from an initial clinical trial at the Austin Hospital in Melbourne, published today in Critical Care, show that sodium ascorbate – a pH-balanced formulation of vitamin C – is effective in treating sepsis. 

Lead investigator Associate Professor Yugeesh Lankadeva said sepsis is notoriously difficult to treat and is often deadly. 

“Sepsis accounts for 35–50 per cent of all hospital deaths. It occurs when the immune system fails to fight off an underlying infection, causing life-threatening falls in blood pressure, multiple organ failure, and death,” Associate Professor Lankadeva said. 

“In our trial at Austin Hospital, patients were given sodium ascorbate into their bloodstream, resulting in promising improvements to multiple organs.” 

Associate Professor Lankadeva, who is research lead of The Florey’s Systems Neuroscience Theme, said the next phase of the $4.9 million government-funded research project will be offered in intensive care units in Adelaide, Melbourne, Perth, Brisbane, Alice Springs and Sydney. 

“We’ll recruit 300 adult septic patients to receive either our formulation or a placebo as well as normal hospital care. The results will help us gather data to determine the effectiveness of our formulation,” Associate Professor Lankadeva said.  

Austin Hospital’s Director of Intensive Care Research, Professor Rinaldo Bellomo, said the previous part of the trial in his department involved 30 adult sepsis patients between October 2020 and November 2022. 

While receiving hospital intensive care, half the patients were randomly allocated to receive sodium ascorbate, and the other half a placebo. 

The study found that patients with sepsis who received the sodium ascorbate treatment: 

• produced more urine, a sign of improved kidney function 

• required less of the clinically used drug, noradrenaline, to restore blood pressure 

• showed signs of improved function in multiple organs. 

“Sepsis is the biggest killer in intensive care units in Australia and worldwide,” Professor Bellomo said. “It often develops so quickly that patients are already critically ill by the time they reach us. A treatment that acts quickly, is safe and highly effective would be an absolute game-changer.” 

Senior Florey researcher on the project, Professor Clive May has spent more than 20 years investigating how sepsis causes organ failure, particularly in the brain and kidneys. 

“Through our showing reduced levels of oxygen in tissues in sepsis, we identified sodium ascorbate as a possible treatment. 

“We’ve seen dramatic results in our work with preclinical studies, where an extremely high dose of sodium ascorbate resulted in full recovery within just three hours, with no side effects. It’s heartening to see all those years of painstaking research pay off with a treatment now within reach for patients,” Professor Clive May said. 

Surviving sepsis: A patient’s perspective  

Long-term Florey staff member Brett Purcell is a consumer representative on the MEGASCORES research program, providing the invaluable perspective of a sepsis survivor. 

"I was rushed to hospital by ambulance in 2011 with a high temperature and delirium. I had the early stages of sepsis. My condition slowly deteriorated and after 12 days I was transferred to a bigger hospital. By then I was in septic shock with a severely infected heart. 

“Six months earlier I had a successful aortic valve replacement; unfortunately the valve had become infected. 

“The surgical team repaired the damage in a six-hour operation. But my condition worsened to critical. My family was told I only had an hour to live as my blood oxygenation levels were not conducive to life. What saved me was the excellent decision making by the surgical team and the ICU intensivists. I was placed on life support on an ECMO machine and given dialysis and my condition rapidly improved. 

“I returned home after almost eight weeks in hospital. 

“I’m so lucky to be alive, and I’m hoping that this new research with sodium ascorbate will provide hospitals with a new life-saving tool, one that’s less invasive, quick and highly effective in fighting sepsis." 

L-R Florey Professor Clive May, Austin Health intensive care specialist Professor Rinaldo Bellomo and Florey Associate Professor Yugeesh Lankadeva have found sodium ascorbate can be used to treat sepsis.

CREDIT

Supplied by The Florey

JOURNAL

Critical Care

DOI

10.1186/s13054-023-04644-x 

METHOD OF RESEARCH

Randomized controlled/clinical trial

SUBJECT OF RESEARCH

People

ARTICLE TITLE

Mega-dose sodium ascorbate: a pilot, single-dose, physiological efect, double-blind, randomized, controlled trial

COI STATEMENT

Yugeesh Lankadeva, Clive May and Rinaldo Bellomo hold a patent for the use of mega dose sodium ascorbate in the treatment of sepsis.

 

HSE neuroscientists use neural network to enhance neurofeedback technology

They succeeded in reducing the delay between a change in brain activity and the corresponding feedback signal by a factor of 50

Peer-Reviewed Publication

NATIONAL RESEARCH UNIVERSITY HIGHER SCHOOL OF ECONOMICS

Researchers from HSE University and the Artificial Intelligence Research Institute (AIRI) have successfully lowered the latency between a change in brain activity and the presentation of the corresponding neurofeedback signal by a factor of 50. The results were obtained by employing a neural network trained in low-latency filtering of brain activity signals from diverse individuals. This approach opens up new prospects for the treatment of attention deficit disorder and epilepsy. A paper with the study findings has been published in Journal of Neural Engineering.

Neurofeedback, a form of biofeedback, has been in use since the 1960s. Its core concept involves individuals receiving objective information about the parameters of their own brain activity, as recorded using an electroencephalogram (EEG), and subsequently learning to regulate their brain waves based on this feedback. For instance, a person can improve their relaxation skills by receiving feedback about the alpha rhythms in their  parietal lobe, since an increase in their intensity typically coincides with the state of relaxation. Neurofeedback technology has a broad range of applications, spanning fromthe treatment of conditions like attention deficit hyperactivity disorder (ADHD), epilepsy, and depression to enhancing stress resilience and athlete training.

However, in practice, not all individuals undergoing neurofeedback training experience substantial improvements—approximately 40% show minimal or no progress. According to HSE researchers, one of the primary reasons for this is the significant delay that occurs between the alteration in brain activity and the presentation of a feedback signal reflecting this change.

"Previously, we discovered that during occipital alpha rhythm training, the frequency of brain activity bursts per unit of time alters, while their duration and amplitude remain constant. The concept behind this training is that individuals can learn to induce a state that leads to an increased frequency of these bursts, for which timely positive reinforcement of such transitions becomes crucial. However, in the majority of systems used today, the feedback signal is delivered with a delay exceeding 500 ms. Under such circumstances, establishing a correlation between the feedback and the corresponding event becomes difficult," says Alexei Ossadtchi, Research Team Leader, Director of the Centre for Bioelectric Interfaces at the HSE University Institute for Cognitive Neuroscience, Head of the Neurointerfaces Group at AIRI.

Decreasing the delay in presenting the feedback signal increases the likelihood of activating the neuroplasticity mechanisms required to attain a lasting effect from the training. In a previous study, all participants in groups experiencing a minimal delay, which was set at 250 ms, successfully increased the frequency of alpha rhythm bursts per unit of time. However, in groups experiencing a delay of approximately 500 ms and higher, only about 60% of participants were able to accomplish the task.

According to the researchers, further reduction of the delay is likely to result in an even more pronounced acceleration in the learning process and to the attainment of long-term training effects. However, the most substantial component of the delay in presenting the feedback signal is linked to fundamental constraints.

The issue lies in Gabor's uncertainty principle. To isolate the rhythm, it is necessary to use the recorded signal values and to observe the signal over a time interval of approximately 200–300 ms. This means that the filtering process—selecting the relevant brain rhythms—takes time and thus delays the signal. The researchers have suggested employing a neural network model of the target signal to accelerate its detection amidst the rest of the brain's activity.

The scientists trained multiple neural networks using extensive datasets of individuals' brain activity, then assessed their stability by introducing noise, and subsequently applied them to data from 25 subjects undergoing alpha-rhythm training. Various architectures were tested, and the Temporal Convolutional Network (TCN) demonstrated the best performance.

"Based on the TCN, a filter was constructed to isolate rhythmic activity, leading to a reduction in the delay of presenting a feedback signal which reflected the instantaneous intensity of the alpha rhythm to just 10 ms. Hence, we have lowered the delay by approximately fifty-fold as compared to the majority of neuroefeedback systems. At the same time, we monitored the transitions of neural populations from the excitation phase to the inhibition phase with virtually no delay," explains Alexei Ossadtchi. 

According to the study authors, their findings can justify a reassessment of the efficacy of neurofeedback in addressing various neurological disorders. Employing the method with reduced delay can substantially increase the proportion of patients who respond positively to this therapy for nervous system dysfunction. Furthermore, this opens up exciting future  possibilities of developing closed-loop brain stimulation paradigms for the treatment of severe neurological disorders by establishing artificial feedback loops that the brain cannot distinguish from its natural feedback mechanisms, thus inducing targeted plastic changes within the brain's neural networks.

---

JOURNAL

Journal of Neural Engineering

DOI

10.1088/1741-2552/acf7f3 

ARTICLE TITLE

Real-time low latency estimation of brain rhythms with deep neural networks

 

MOBIOS+: An extensive database for advancing Mindanao Island’s biodiversity

Peer-Reviewed Publication

PENSOFT PUBLISHERS

 

IMAGE: 

TEAM MEMBERS OF THE MOBIOS+ CONSORTIUM CURATING THE DATASET.

view more 

CREDIT: TANALGO ET AL.

The Philippine Archipelago, with more than 7,100 islands, has one of the highest levels of endemism globally and is a hotspot for biodiversity conservation. Mindanao, the second largest group of islands in the country, is a treasure trove of terrestrial species, boasting one of the highest densities of unique flora and fauna on the planet. However, despite its ecological significance, comprehensive biodiversity records and data for the region have remained inaccessible until now.

The Mindanao Open Biodiversity Information (MOBIOS+) database aims to bridge these critical data gaps by compiling biodiversity information from the 21st century. This monumental undertaking seeks to enhance our understanding of Mindanao's biodiversity trends, while establishing a database that is openly accessible to researchers and conservationists worldwide.

MOBIOS+ is the first of its kind and, currently, the most comprehensive attempt to create a consolidated database for the biodiversity of Mindanao based on publicly available literature. With a vast collection of biodiversity data, this database will be an invaluable resource to advance regional biodiversity research and analysis.

“It will further facilitate the identification of species and areas that require immediate conservation prioritisation and action, addressing the urgent challenges posed by our rapidly changing planet,” the researchers behind the project write in their data paper, published in the open-access, peer-reviewed Biodiversity Data Journal.

The MOBIOS+ database, available through the Global Biodiversity Information Facility (GBIF) platform, currently comprises an impressive 12,813 georeferenced specimen occurrences representing 1,907 unique taxa. These span across ten animal classes inhabiting terrestrial and freshwater environments within the Mindanao faunal region. The project aims to continuously update the species database, complementing on-ground biodiversity efforts in Mindanao.

Associate Professor Krizler Tanalgo of the Ecology and Conservation Research Laboratory at the University of Southern Mindanao, the project leader behind MOBIOS+, shared his thoughts on this initiative, saying:

“We aim to democratise biodiversity information, making it readily available to researchers, policymakers, and conservation biologists. By doing so, we hope to facilitate well-informed decisions to address pressing environmental challenges, with a particular focus on the often underrepresented Mindanao region, which tends to receive limited attention in terms of research and funding.”

“The MOBIOS+ database is not only a testament to the dedication of the scientific community, but also a beacon of hope for the future of biodiversity conservation in Mindanao and beyond. It will support researchers and conservationists in identifying species and areas that require immediate prioritisation and action, safeguarding the unique and fragile ecosystems of this extraordinary region.”

The Biodiversity Community Integrated Knowledge Library (BiCIKL) project, funded by the European Union Horizon 2020 Research and Innovation Action under grant agreement No 101007492, has supported the publication of this work. The work is part of a special collection supported by the project and looking to demonstrate the advantages and novel approaches in accessing and (re-)using linked biodiversity data.

 

 

Original source:
Tanalgo KC, Dela Cruz KC, Agduma AR, Respicio JMV, Abdullah SS, Alvaro-Ele RJ, Hilario-Husain BA, Manampan-Rubio M, Murray SA, Casim LF, Pantog AMM, Balase SMP, Abdulkasan RMA, Aguirre CAS, Banto NL, Broncate SMM, Dimacaling AD, Fabrero GVN, Lidasan AK, Lingcob AA, Millondaga AM, Panilla KFL, Sinadjan CQM, Unte ND (2023) The MOBIOS+: A FAIR (Findable, Accessible, Interoperable and Reusable) database for Mindanao's terrestrial biodiversity. Biodiversity Data Journal 11: e110016. https://doi.org/10.3897/BDJ.11.e110016

---

JOURNAL

Biodiversity Data Journal

DOI

10.3897/BDJ.11.e110016 

ARTICLE TITLE

The MOBIOS+: A FAIR (Findable, Accessible, Interoperable and Reusable) database for Mindanao's terrestrial biodiversity

 

Widely tuneable terahertz lasers boost photo-induced superconductivity in K3C60

Peer-Reviewed Publication

MAX PLANCK INSTITUTE FOR THE STRUCTURE AND DYNAMICS OF MATTER

 

IMAGE: 

THE OPTICAL SETUP WHICH WAS USED TO GENERATE THE LOW FREQUENCY LASER PULSES WHICH RESONANTLY EXCITED THE SAMPLE.

view more 

CREDIT: J. WITT, MPSD

Researchers at the Max Planck Institute for the Structure and Dynamics of Matter (MPSD) in Hamburg, Germany, have long been exploring the effect of using tailored laser drives to manipulate the properties of quantum materials away from equilibrium. One of the most striking demonstrations of these physics has been in unconventional superconductors, where signatures of enhanced electronic coherences and super-transport have been documented in the resulting non-equilibrium states. However, these phenomena have not yet been systematically studied or optimized, primarily due to the complexity of the experiments. Technological applications are therefore still far removed from reality.

In a recent experiment, this same group of researchers discovered a far more efficient way to create a previously observed metastable, superconducting-like state in K3C60 using laser light. The work by the Cavalleri group has appeared in Nature Physics.

The researchers showed that, when tuning the laser light to a specific low frequency resonance, far less powerful light pulses could induce the same effect at much higher temperatures. Laser technology developed at the Institute was key to this work. By tuning the light source to 10 THz, a lower frequency than previously possible, the team successfully recreated the long-lived superconducting-like state in the fullerene-based material while reducing the pulse intensity by a factor of 100. This light-induced state was directly observed to persist at room temperature for 100 picoseconds, but is predicted to have a lifetime of at least 0.5 nanoseconds (a nanosecond is a billionth of a second, a picosecond a trillionth).

Their findings shed new light on the underlying microscopic mechanism in photo-induced superconductivity, says lead author Edward Rowe, a PhD student in the Cavalleri group: “The identification of the resonance frequency will allow theorists to understand which excitations are actually important, since there is currently no widely accepted theoretical explanation of this effect in K3C60.”

Rowe envisages that a light source with a higher repetition rate at the 10 THz frequency could help sustain the metastable state for longer: “If we could deliver each new pulse before the sample returns to its non-superconducting equilibrium state, it may be possible to sustain the superconducting-like state continuously.”

“These experiments are a very nice demonstration of how suitable advances in technology can make applicable many phenomena that are so far not practical,” says MPSD Director Andrea Cavalleri, who sees a two-decade long effort in exploring these effects converging towards future technologies. “It is also clear that a crucial bottleneck to be addressed is the type and availability of laser sources, which should go hand in hand with these studies to bring the field forward.”

The research was carried out at the MPSD, in the Center for Free-Electron Laser Science (CFEL) in Hamburg. It was supported by the DFG (German Research Foundation) via the Cluster of Excellence CUI : Advanced Imaging of Matter. The K3C60 samples were prepared at the Università degli Studi di Parma, Italy.

---

JOURNAL

Nature Physics

METHOD OF RESEARCH

Experimental study

ARTICLE TITLE

Resonant enhancement of photo-induced superconductivity in K3C60

AI just got 100-fold more energy efficient

Nanoelectronic device performs real-time AI classification without relying on the cloud

Peer-Reviewed Publication

NORTHWESTERN UNIVERSITY

• 
  
• 
  
• 
  
• AI is so energy hungry that most data analysis must be performed in the cloud
• New energy-efficient device enables AI tasks to be performed within wearables
• This allows real-time analysis and diagnostics for faster medical interventions
• Researchers tested the device by classifying 10,000 electrocardiogram samples
• The device successfully identified six types of heart beats with 95% accuracy

EVANSTON, Ill. — Forget the cloud.

Northwestern University engineers have developed a new nanoelectronic device that can perform accurate machine-learning classification tasks in the most energy-efficient manner yet. Using 100-fold less energy than current technologies, the device can crunch large amounts of data and perform artificial intelligence (AI) tasks in real time without beaming data to the cloud for analysis.

With its tiny footprint, ultra-low power consumption and lack of lag time to receive analyses, the device is ideal for direct incorporation into wearable electronics (like smart watches and fitness trackers) for real-time data processing and near-instant diagnostics.

To test the concept, engineers used the device to classify large amounts of information from publicly available electrocardiogram (ECG) datasets. Not only could the device efficiently and correctly identify an irregular heartbeat, it also was able to determine the arrhythmia subtype from among six different categories with near 95% accuracy.

The research will be published on Oct. 12 in the journal Nature Electronics.

“Today, most sensors collect data and then send it to the cloud, where the analysis occurs on energy-hungry servers before the results are finally sent back to the user,” said Northwestern’s Mark C. Hersam, the study’s senior author. “This approach is incredibly expensive, consumes significant energy and adds a time delay. Our device is so energy efficient that it can be deployed directly in wearable electronics for real-time detection and data processing, enabling more rapid intervention for health emergencies.”

A nanotechnology expert, Hersam is Walter P. Murphy Professor of Materials Science and Engineering at Northwestern’s McCormick School of Engineering. He also is chair of the Department of Materials Science and Engineering, director of the Materials Research Science and Engineering Center and member of the International Institute of Nanotechnology. Hersam co-led the research with Han Wang, a professor at the University of Southern California, and Vinod Sangwan, a research assistant professor at Northwestern.

Before machine-learning tools can analyze new data, these tools must first accurately and reliably sort training data into various categories. For example, if a tool is sorting photos by color, then it needs to recognize which photos are red, yellow or blue in order to accurately classify them. An easy chore for a human, yes, but a complicated — and energy-hungry — job for a machine.

For current silicon-based technologies to categorize data from large sets like ECGs, it takes more than 100 transistors — each requiring its own energy to run. But Northwestern’s nanoelectronic device can perform the same machine-learning classification with just two devices. By reducing the number of devices, the researchers drastically reduced power consumption and developed a much smaller device that can be integrated into a standard wearable gadget.

The secret behind the novel device is its unprecedented tunability, which arises from a mix of materials. While traditional technologies use silicon, the researchers constructed the miniaturized transistors from two-dimensional molybdenum disulfide and one-dimensional carbon nanotubes. So instead of needing many silicon transistors — one for each step of data processing — the reconfigurable transistors are dynamic enough to switch among various steps.

“The integration of two disparate materials into one device allows us to strongly modulate the current flow with applied voltages, enabling dynamic reconfigurability,” Hersam said. “Having a high degree of tunability in a single device allows us to perform sophisticated classification algorithms with a small footprint and low energy consumption.”

To test the device, the researchers looked to publicly available medical datasets. They first trained the device to interpret data from ECGs, a task that typically requires significant time from trained health care workers. Then, they asked the device to classify six types of heart beats: normal, atrial premature beat, premature ventricular contraction, paced beat, left bundle branch block beat and right bundle branch block beat.

The nanoelectronic device was able to identify accurately each arrhythmia type out of 10,000 ECG samples. By bypassing the need to send data to the cloud, the device not only saves critical time for a patient but also protects privacy.

“Every time data are passed around, it increases the likelihood of the data being stolen,” Hersam said. “If personal health data is processed locally — such as on your wrist in your watch — that presents a much lower security risk. In this manner, our device improves privacy and reduces the risk of a breach.”

Hersam imagines that, eventually, these nanoelectronic devices could be incorporated into everyday wearables, personalized to each user’s health profile for real-time applications. They would enable people to make the most of the data they already collect without sapping power.

“Artificial intelligence tools are consuming an increasing fraction of the power grid,” Hersam said. “It is an unsustainable path if we continue relying on conventional computer hardware.”

The study, “Reconfigurable mixed-kernel heterojunction transistors for personalized support vector machine classification,” was supported by the U.S. Department of Energy, National Science Foundation and Army Research Office.

---

JOURNAL

Nature Electronics

DOI

10.1038/s41928-023-01042-7 

METHOD OF RESEARCH

Experimental study

SUBJECT OF RESEARCH

Not applicable

ARTICLE TITLE

Reconfigurable mixed-kernel heterojunction transistors for personalized support vector machine classification

ARTICLE PUBLICATION DATE

12-Oct-2023

AI researchers expose critical vulnerabilities within major LLMs

Computer scientists from the AI security start-up Mindgard and Lancaster University in the UK have demonstrated that chunks of these LLMs can be copied in less than a week for as little as $50, and the information gained can be used to launch targeted attacks.

Reports and Proceedings

LANCASTER UNIVERSITY

Large Language Models (LLMs) such as ChatGPT and Bard have taken the world by storm this year, with companies investing millions to develop these AI tools, and some leading AI chatbots being valued in the billions.

 

These LLMs, which are increasingly used within AI chatbots, scrape the entire Internet of information to learn and to inform answers that they provide to user-specified requests, known as ‘prompts’.

 

However, computer scientists from the AI security start-up Mindgard and Lancaster University in the UK have demonstrated that chunks of these LLMs can be copied in less than a week for as little as $50, and the information gained can be used to launch targeted attacks.

 

The researchers warn that attackers exploiting these vulnerabilities could reveal private confidential information, bypass guardrails, provide incorrect answers, or stage further targeted attacks.

 

Detailed in a new paper to be presented at CAMLIS 2023 (Conference on Applied Machine Learning for Information Security) the researchers show that it is possible to copy important aspects of existing LLMs cheaply, and they demonstrate evidence of vulnerabilities being transferred between different models.

 

This attack, termed  ‘model leeching’, works by talking to LLMs in such a way – asking it a set of targeted prompts – so that the LLMs elicit insightful information giving away how the model works.

 

The research team, which focused their study on ChatGPT-3.5-Turbo, then used this knowledge to create their own copy model, which was 100 times smaller but replicated key aspects of the LLM.

 

The researchers were then able to use this model copy as a testing ground to work out how to exploit vulnerabilities in ChatGPT without detection. They were then able to use the knowledge gleaned from their model to attack vulnerabilities in ChatGPT with an 11% increased success rate.

 

Dr Peter Garraghan of Lancaster University, CEO of Mindgard, and Principal Investigator on the research, said: “What we discovered is scientifically fascinating, but extremely worrying. This is among the very first works to empirically demonstrate that security vulnerabilities can be successfully transferred between closed source and open source Machine Learning models, which is extremely concerning given how much industry relies on publicly available Machine Learning models hosted in places such as HuggingFace.”

 

The researchers say their work highlights that although these powerful digital AI technologies have clear uses, there exist hidden weaknesses, and there may even be common vulnerabilities across models.

 

Businesses across industry are currently or preparing to invest billions in creating their own LLMs to undertake a wide range of tasks such as smart assistants. Financial services and large enterprises are adopting these technologies but researchers say that these vulnerabilities should be a major concern for all businesses that are planning to build or use third party LLMs.

 

Dr Garraghan said: “While LLM technology is potentially transformative, businesses and scientists alike will have to think very carefully on understanding and measuring the cyber risks associated with adopting and deploying LLMs.”

 

The paper will be presented at CAMLIS 2023 in Arlington, Virginia USA which is held on October 19 and 20.

 

The paper’s authors are Lewis Birch, William Hackett, Stefan Trawicki, and Neeraj Suri of Lancaster University, and Peter Garraghan of Lancaster University and Mindgard.

 

ENDS

 

Letting go of an extra weight to control sleeping sickness

Peer-Reviewed Publication

INSTITUTO DE MEDICINA MOLECULAR

 

IMAGE: 

LUÍSA FIGUEIREDO, LEADER OF THE LAB WHERE THE STUDY WAS PERFORMED. 

view more 

CREDIT: GONÇALO RIBEIRO

A new study led by Luísa Figueiredo, group leader at the Instituto de Medicina Molecular João Lobo Antunes (iMM; Portugal), and published today in the scientific journal Nature Microbiology* found a new strategy by the host to cope with Trypanosoma brucei infection. Trypanosoma brucei is the parasite that causes sleeping sickness in humans, and nagana in cattle, which remain a public health concern and a major economic burden in Sub-Saharan Africa. Understanding the response of the infected organism during disease progression paves the way for the development of new treatment strategies.

Upon the bite of an infected tsetse fly, the parasites colonise intersticial spaces in the mammalian hosts, leading to death if the infection is left untreated. “Some years ago, our laboratory found that the parasite that causes sleeping sickness, Trypanosoma brucei, accumulates in disproportionately high numbers in the adipose tissue. We found this very intriguing, and directed our research to the understanding of how adipose tissue colonisation affects disease progression”, introduces Luísa Figueiredo, leader of the lab where the study was carried out. “The accumulation of the parasites in these tissues is accompanied by weight loss. Now, we found that the weight loss is largely due to the loss of adipose tissue by lipolysis, a metabolic process that breaks triacylglycerols into their constituent molecules, including free fatty acids”, continues Luísa Figueiredo. The loss of adipose tissue can occur due to many different signals. “We found that in this case, the lipolysis is in large part induced by the host immune response to the infection”, says Henrique Machado, first author of the study, explaining: “We infected immunocompromised mice and mice with impaired lipolysis, and saw that, in both cases, there is less loss of body weight despite having higher number of parasites”.

Then, the authors set out to understand if the lipolysis contributes to the progression of the disease. “We saw that the loss of adipose tissue through lipolysis prolongs the life of the infected mice. This, adding to the fact that the lipolysis exerts a control on the number of parasites, indicates that this response has a protective effect, somehow. Interestingly, we discovered that this protective effect is due to the release of the free fatty acids, during lipolysis, that are toxic for the parasites and lead to parasite death”, explains Luísa Figueiredo, saying: “It’s as if the host body sacrifices itself, by losing fat mass, to kill the parasites”.

These findings can open a completely new line of treatments for sleeping sickness and nagana targeting the infection in the adipose tissue, since the relevance of this tissue was previously unknown.

 

*Henrique Machado, Peter Hofer, Rudolf Zechner, Terry K. Smith, Luísa M. Figueiredo. 2023. Adipocyte lipolysis protects the host against Trypanosoma brucei infection. Nature Microbiology. DOI: 10.1038/s41564-023-01496-7.

 

This work was developed at iMM in collaboration with the University of Graz, Austria, the BioTechMed-Graz, Austria, and the University of St Andrews, United Kingdom. This work was supported by the European Research Council, the Portuguese Foundation for Science and Technology, the Austrian Fonds zur Förderung der Wissenschaftlichen Forschung, and the Foundation Loius Jeantet. 
 

---

JOURNAL

Nature Microbiology

DOI

10.1038/s41564-023-01496-7 

METHOD OF RESEARCH

Experimental study

SUBJECT OF RESEARCH

Animals

ARTICLE TITLE

Adipocyte lipolysis protects the host against Trypanosoma brucei infection.

ARTICLE PUBLICATION DATE

12-Oct-2023

 

How do tax proposals affect cancer health of tobacco users based on income, education?

Grant and Award Announcement

VIRGINIA TECH

 

IMAGE: 

RESEARCH ASSISTANT PROFESSOR ROBERTA FREITAS-LEMOS OF THE FRALIN BIOMEDICAL RESEARCH INSTITUTE AT VTC RECEIVED A FIVE-YEAR, $680,000 AWARD FROM THE NATIONAL CANCER INSTITUTE TO STUDY THE INTERPLAY OF CANCER, TOBACCO POLICIES, AND HEALTH DISPARITIES.

view more 

CREDIT: PHOTO BY CLAYTON METZ FOR VIRGINIA TECH.

Tobacco is the leading cause of preventable death in the United States, according to the Centers for Disease Control and Prevention, and cigarette smoking causes three in 10 of all cancer deaths. Smoking also accounts for more than 30 percent of the difference in life expectancy among different socioeconomic groups.

Roberta Freitas-Lemos, research assistant professor at the Fralin Biomedical Research Institute at VTC, recently received a career development award to explore the ways in which nicotine tax policies can influence health disparities. The award of more than $680,000 over five years from the National Cancer Institute of the National Institutes of Health is designed to promote career development and mentored training of cancer researchers from underrepresented groups. 

It will provide Freitas-Lemos resources to launch independent, cancer-related research on eliminating tobacco use and its unequal harms. 

“My previous work has focused on people living in vulnerable situations and forecasting the effects of regulatory policies on tobacco and nicotine consumption among different users,” she said.

Freitas-Lemos studied psychology at the Pontifical Catholic University in São Paulo, Brazil, before earning a doctorate from the University of Brasilia. She joined the Fralin Biomedical Research Institute as postdoctoral fellow in 2019.

Taxes are widely used to reduce tobacco use and can by extension reduce the incidence of cancer. But they can exacerbate socioeconomic disparities. Freitas-Lemos is testing a tax proposal designed to reduce tobacco dependence and tobacco-related disparities in cigarette smokers from diverse socioeconomic backgrounds. 

She plans to develop a new tax proposal based on the abuse liability of tobacco products and investigate its effects on purchase behavior using the Experimental Tobacco Marketplace, a tool developed at the Fralin Biomedical Research Institute. Her proposal uses an abuse liability assessment model, which predicts the likelihood of tobacco addiction or harmful use.

Warren Bickel, professor at the Fralin Biomedical Research Institute at VTC and Freitas-Lemos’ primary mentor on the project, developed the marketplace to explore the effect of tax and regulatory policy on nicotine purchases. Bickel is also director of the institute's Addiction Recovery Research Center.

Freitas-Lemos will recruit a diverse group of tobacco users who will use an experimental account to purchase tobacco and replacement therapy products through the marketplace. By adjusting the product mix and pricing, scientists can better predict purchase behavior. 

“My long-term research goal is to become an independent researcher investigating the differential impact of policies on tobacco initiation, use and cessation among individuals who experience tobacco-related cancer disparities,” Freitas-Lemos said.