Wednesday, October 28, 2026

 

A portable device for accurately and rapidly detecting PFAS in water



Inexpensive electrodes combined with new ‘synthetic antibodies’ detect trace levels of “forever chemicals” in water, opening a promising route to portable environmental monitoring




Okinawa Institute of Science and Technology (OIST) Graduate University

PFAS sensor prototype

image: 

The prototypical device. An electrode (the three dark prongs on the white strip) is screen-printed on fabric strips, the same way that an illustration might be printed onto a T-shirt, and coated with molecularly imprinted polymers (MIPs) that bind specifically to either PFOS or PFOA. This strip is then inserted into a potentiostat (the black dongle, which measures electrical current), which is plugged directly into a smartphone that can take the PFAS measurement.  

view more 

Credit: Andrew Scott & Jeff Prine/OIST






Per- and polyfluoroalkyl substances (PFAS) are synthetic chemicals known for their exceptional resistance to degradation, earning them the nickname “forever chemicals.” These persistent substances can accumulate in water, soil, and living organisms, raising growing concerns about their potential impacts on human health and the environment.

Detecting PFAS contamination is a critical first step toward addressing these risks. However, conventional PFAS analysis typically requires collecting environmental samples and transporting them to specialized laboratories equipped with advanced mass spectrometry instruments. Although highly sensitive and accurate, these methods can be costly and time-consuming.

Researchers from the Micro/Bio/Nanofluidics Unit at the Okinawa Institute of Science and Technology (OIST) have now created a reliable, easy-to-mass-produce platform that enables regulators and other environmental actors to detect PFAS on the spot without the need for an external lab. “Our goal is to develop a sensitive, low-cost technology that can eventually be used outside specialized laboratories,” says Dr. Mark Sullivan. “We have demonstrated that it works for PFOA and PFOS, and the same platform could potentially be adapted to detect other PFAS compounds.”

The new device is described in Biosensors and Bioelectronics.

Catching PFAS with synthetic antibodies and electrical currents

The researchers have designed a new class of molecularly imprinted polymers, or MIPs, that are perfectly tailored to fit PFAS. The materials are formed around the target molecule — in this case, PFAS — creating a binding site. The target molecule is then washed away, leaving behind a material that recognizes and captures the target again, like a lock that is perfectly shaped for a key. This is also why MIPs are often called ‘synthetic antibodies.’

Their device detects PFAS by measuring changes in electrical current as target molecules bind to a sensor coated with PFAS-specific MIPs. The greater the change in electrical current, the higher the PFAS concentration is in the water sample. Because the measurement requires no additional chemical reagents, this approach could provide a simpler, lower-cost solution for portable, on-site water monitoring.

The platform is compact, low-cost, and designed for easy manufacturing. For the prototype, the researchers screen-printed 100 disposable electrodes onto repurposed fabric in just a few hours. These were then coated with PFAS-specific MIPs and connected to commercially available potentiostats, which measure electrical signals and can plug directly into a smartphone or laptop. Furthermore, the MIP manufacturing method is modular, meaning the target molecule can be replaced with any member of the PFAS family; for this prototype, the researchers chose the two most common PFAS chemicals with documented health risks: PFOS and PFOA.

Strengthening the global environmental immune defense against forever chemicals

PFAS have proven to be an insidious problem for the environment and health due to two key factors: their exceptional resistance to heat, water, oil, and environmental degradation makes them both useful and extremely difficult to break down; and their diversity, with the PFAS family encompassing more than 10,000 similar chemicals, of which 50-150 members are regulated across the globe.

The team’s MIP platform’s modularity makes it relatively simple to adapt it to different PFAS molecules and, more importantly, lowers the barrier to addressing PFAS contamination going forward. Dr. Perrine Lasserre continues:

“It shows that it’s possible to make a detection system that empowers local agencies or even families to make an informed decision about whether and how to address potential PFAS contamination — such as alerting municipal authorities to contamination at specific sites, or whether to invest in a water filter. Whatever we can achieve with this system is going to be beneficial.”

“Accurate and accessible detection is an important step toward addressing PFAS contamination,” says Professor Amy Shen. “Our study demonstrates how synthetic molecular recognition can be combined with low-cost electrochemical devices to detect extremely small concentrations of these persistent chemicals. While further development and field validation are needed, we see considerable potential for making environmental monitoring more accessible and widely available.”

 

GenAI and travel planning: should AI agents build our itineraries?



A new Edith Cowan University (ECU) study of more than 400 Australian travellers has set out to discover what the most influential drivers are behind our decisions to delegate travel planning to AI, and what our threshold is when it comes to trust




Edith Cowan University

Professor Sanjit Roy

image: 

Professor Sanjit Roy, Edith Cowan University.

view more 

Credit: Edith Cowan University






Whether you’re a seasoned tourist or an emerging explorer the decision to delegate travel planning to artificial intelligence (AI) agents is becoming the norm for many travellers.

A new Edith Cowan University (ECU) study of more than 400 Australian travellers has set out to discover what the most influential drivers are behind our decisions to delegate travel planning to AI, and what our threshold is when it comes to trust.

ECU Professor of Marketing and Service Science, Sanjit Roy from ECU’s School of Business and Law said the use of responsible AI in tourism is critically important – not only for planning great experiences but ensuring travellers are prepared when things go wrong.

A role to play in tourism

“Responsible use of AI not only gives more certainty that trips will be smooth sailing, but holidaymakers need to ensure they aren’t left out of pocket or on the backfoot when something goes wrong,” Professor Roy said.

The need to safeguard against the trip of a lifetime turning into a tour of misery lane means travellers must be willing to become AI-savvy. Conversely, AI needs to do the right thing by the user.

“Generative Artificial Intelligence (GenAI) is becoming an integral part of travel planning and revolutionising travel through automated planning systems,” he said.

“Customers need to be able to trust AI agents to comfortably outsource building their itineraries.”

A recent survey of adult travellers revealed 28.8 per cent of Australians use GenAI to find deals, check out destinations and create their own travel itineraries. Among Gen Z travellers, Australian Government research puts the figure at 30 per cent.

This fundamental transformation is changing the definition of travel decision delegation.

Accountability key

The ECU study analysed data collected from 421 Australian travellers, highlighting five key responsible AI characteristics travellers used to evaluate information produced from GenAI.

“Our findings tell us that travellers consider using GenAI for travel planning based on five characteristics: reliability, accountability, trustworthiness, fairness and transparency. In particular, reliability and accountability play a crucial role in planning travel experiences,” Professor Roy explained.

“Travellers were more likely to use AI when they perceived it as reliable and believed its providers could be held accountable for its recommendations and for addressing problems when they arise.

“Accountability of AI plays a significant role in building trust among travellers.”

The transparency paradox

Interestingly, too much information steered travellers away from using GenAI.

“We found that excessive transparency reduces our willingness to delegate travel planning to GenAI – meaning the mental burden of processing, cross-checking and translating information is not worth it,” Professor Roy said.

The Australian government’s Smartraveller website warns AI can provide outdated transport schedules, incorrect visa information, and recommendations for hotels that have closed. Relying on incorrect information could ruin part of your dream vacation.

GenAI not a lost cause

Whether you use ChatGPT, Claude, Gemini, Mindtrip, Layla, Wanderlog or other – Professor Roy said vacationers shouldn’t dismiss AI as a useful tool to inspire travel itineraries.

“When AI is reliable, trustworthy, honest and doesn’t overload travellers with unnecessary information, they are more likely to delegate travel planning decisions to AI agents,” he said.

When planning an overseas vacation, travellers should fact-check the information provided, know who is accountable if something goes wrong and not assume that more information means the source is highly trustworthy.

“We know that not all those who wander are lost. But not all those who wonder about using AI to plan their next trip will find themselves lost either. You may save precious time planning an incredible travel experience with the help of GenAI.”

Decision delegation to GenAI agents in travel planning: Responsible AI signals, delegation levels, and the transparency paradox is published in Tourism Management Perspectives.

- ends -   

Media contact:

Hayley Butler, (08) 6304 5575, h.butler@ecu.edu.au

or

ECU Corporate Relations, (08) 6304 2222, pr@ecu.edu.au

Sign up to get the latest research news and media releases delivered to your inbox.

 

 

3 out of 4 responses to UK smoke-free generation consultation were AI bot-generated



And overwhelmingly targeted at opposing restrictions on disposable vape sales; scale and consistency of responses suggest a coordinated campaign




BMJ Group






Three quarters of all the responses to the UK’s public consultation in 2023 on creating a smoke-free generation and restricting disposable vape sales were generated by AI bots, finds an analysis of publicly available data and Freedom of Information requests, published online in the journal Tobacco Control.

 

These responses were overwhelmingly targeted at opposing restrictions on disposable vape sales, the scale and directional consistency of which suggest a coordinated campaign, say the researchers.

 

AI bots have been used to spread health misinformation on social media platforms, point out the researchers. But while public consultations are increasingly used to inform health regulation, they are susceptible to automated interference, they add.

 

The researchers therefore wanted to find out if the 2023 public consultation on government proposals to create a smoke-free generation and curb vaping among young people had been subject to automated interference.

 

They drew on published consultation reports and Freedom of Information requests to the Department of Health and Social Care.

 

In all, 118,756 responses were submitted to the consultation, three quarters of which (90,835, 76%), the Department of Health and Social Care identified as fraudulent using bot detection methods.

 

All the bot-generated responses targeted a single multiple choice question opposing proposed restrictions on disposable vape sales. The same number reported learning of the consultation via social media.

 

But after the bot responses were excluded, 77% of the valid respondents supported restrictions on disposable vape sales—the opposite of what the unfiltered dataset would have indicated.

 

“This appears to be the first documented instance of large-scale automated interference in a formal public health policy consultation,” write the researchers.

 

“While it is not possible to ascertain the commercial origins of such a campaign, transnational tobacco companies have sought to swamp earlier public consultations with favourable responses, using mass recruitment campaigns and postcard submissions,” they explain.

 

“The scale and directional consistency of bot responses [in this instance] suggest a coordinated campaign, consistent with prior evidence of tobacco industry efforts to manipulate public consultations through mass coordinated submissions,” they add.

 

The UK’s Tobacco and Vapes Act received Royal Assent on 29 April 2026, making it illegal to ever sell tobacco products to anyone born on or after 1 January 2009, and restricting the supply and sale of disposable vapes.

 

The researchers acknowledge several limitations to their findings. For example, they couldn’t independently verify the bot detection methods used by the Department of Health and Social Care as this information isn’t publicly available, and the metadata and demographic information on the respondents was deemed exempt under Freedom of Information legislation.

 

The findings may not be applicable to other policy topics either, and on the basis of the information received, it’s impossible to attribute source, intent, or commercial sponsorship,  emphasise the researchers.

 

But they conclude: “As bot sophistication increases and public consultations expand across regulatory contexts, strengthening detection capacity and governance safeguards will be essential to maintaining the integrity of public health policymaking, particularly in lower-resource settings where screening infrastructure may be absent.”

 

Promising AI-based approach to early detection of suicide-related behavior at emergency rooms





University of Johannesburg

AI-model flags suicide-related behaviour per 1000 ER visits

image: 

Table showing true/false positives/negatives per 1000 visits, based on the AI confusion matrix in Figure 3 of the research study.

Researchers from the University of Johannesburg (UJ) have published a promising AI-based approach for early detection of people at risk of suicide-related behaviour (SRB) visiting hospital Emergency Rooms (ERs) in the journal Scientific Reports. (https://doi.org/10.1038/s41598-026-64073-y )

The study describes using an AI algorithm, routine patient intake data, and light-weight computing to estimate SRB. The model is intended to assist ER staff, rather than replace them, during very busy periods at resource-constrained hospitals.

view more 

Credit: Therese van Wyk, University of Johannesburg (UJ)




Researchers from the University of Johannesburg (UJ) have published a promising approach to early detection of people at risk of suicide-related behaviour (SRB) in the journal Scientific Reports.

The study describes using an AI algorithm, routine patient intake data, and light-weight computing to estimate SRB risk following admission to a hospital emergency room (ER). The AI model was built with historical open-source patient data from the USA. It is intended to assist ER staff rather than replace them.

Resource constraints in South Africa are a major reason to develop an AI model such as this, says lead researcher Mr Tsholofelo Mokheleli.

Late at night and over weekends, some South African hospital emergency rooms can see one ER staff member with 30 or more patients to triage, he adds.

Mokheleli is an assistant lecturer at the Department of Applied Information Systems within the UJ College of Business and Economics (CBE). The study forms part of his PhD research.

During busy ER shifts, an objective, consistent measure to help staff rapidly screen patients could be helpful for early detection of SRB risk.

“The AI model will flag that a patient is at risk and provide reasons. A health care professional still needs to take the final decision however, the clinician must do a severity scaling to determine if the person is really suicidal or not,” says Mokheleli.

Existing SRB screening is resource intensive

The current two screening methods for SRB used in South African ERs require much time and effort from clinicians, says Prof Nompumelelo Ndlovu.

Prof Ndlovu is an Associate Professor in the Department of Nursing, within the Faculty of Health Sciences at the University of Johannesburg. She is a co-author of the study and one of the co-supervisors of Mokheleli’s PhD research.

When anyone first arrives at an ER, the first priority is to stabilize them medically, she adds.

Even for conscious patients admitted to the ER, the standard mini-mental status examination for SRB risk can take 15 minutes and varies between patients, she says. A full mental status examination and history taking for one patient at high risk of SRB could consume a day or more from a mental health professional at the hospital.

“Not everyone who goes into a medical institution has mental health issues. What we are trying to do, is to find those people who do not even say it out loud,” says Mokheleli.

Adds Prof Ndlovu: “Some people do not realize they are at risk of SRB. People may not be aware of what is happening with them unless they speak to somebody. But it is not an easy topic to speak about. There is so much stigma. So how safe do you feel to tell the next person about what is happening with you?”

Routine admissions data, no new ER workload

The study’s AI model identified patient age, pain, and a prior mood disorder recorded at the same hospital as the major reasons for flagging a patient for SRB risk. Vital signs such as temperature and blood pressure follow on that, with prior substance-use and anxiety diagnoses playing smaller roles.

“The USA patient admission information we used to build this model is similar to what you would get in South Africa,” says Mokheleli.

“It includes demographics, age, gender, vital signs, temperature, heart rate, respiratory rate, oxygen saturations, and systolic and diastolic blood pressure. From that information, we calculate physiological features like the patient’s shock index and pulse pressure,” he adds.

“Another aspect is the patient’s prior history at the same hospital. The model checks if the patient has been at the same hospital before and what the problem was. If there was some kind of substance abuse, anxiety or depression documented with ICD-9 or ICD-10 codes on a previous visit, then the model takes that into account. Night and weekend admissions are also included, as well as notes from medical staff about self-harm on previous visits.”

Local records, local patient data governance

Because the AI model is built on existing electronic patient admissions data, ER staff would not need to do additional data capture for future system implementations. However, in hospitals with paper-based records, the requirement for digitised patient records in a database would be a barrier.

A major reason for building the AI model on routine patient admissions data, is data governance, says co-author and Mokheleli’s PhD main supervisor, Prof Patrick Ndayizigamiye. Prof Ndayizigamiye is a visiting Associate Professor in the Centre of Applied Data Science, at the UJ College of Business and Economics.

“The question is always ‘who owns the confidential patient data and where is it stored?’ Routine admission data belongs to the patient, and it is stored within the hospital’s systems already,” he adds. “As opposed to asking a cloud storage vendor to put the data in their server, which generates a lot of data governance issues.”

False negatives and false positives

With this AI model, a number of factors influence how many patients it will flag correctly for SRB risk. The factors include the size of the dataset; the dynamics of the patient population at the hospital from which data was selected; how few patient records in the training dataset recorded SRB risk; the value of the SRB risk decision threshold; and the AI algorithm selected for analysis.

The same factors also influence how many patients truly at risk for SRB the model misses.

Per 1000 emergency department visits, using the 41,202-patient dataset of the study, the model correctly predicted 866 patients as not at risk of a documented SRB outcome in the 30 days after admission to ER (true negatives).

Per 1000 visits, the model incorrectly predicted or missed 13 cases as not at risk of SRB, although an SRB-related outcome was documented within the 30 days after admission to ER (false negatives).

Per 1000 visits, the model flagged 121 patients as potentially at risk of SRB, and requiring further screening by clinicians (true positives plus false positives), a resource-intensive task.

Of the 121 visits per 1000 the model predicted SRB risk for, 23 had a documented SRB-related outcome that was recorded within the 30 days following the ER admission (true positives).

Of the 121 visits per 1000 flagged for SRB risk, 98 flagged cases did not have any SRB-related outcome documented with ICD-9 or ICD-10 codes in the same hospital’s records, in the 30 days after admission to the ER (false positives).

Source of patient data

Finding a suitable digital patient admissions dataset in South Africa proved difficult, says Prof Ndlovu. To solve this, the study is based on an open dataset from a single hospital in the USA, extracted from the MIMIC-IV resource. The hospital had a mainly White and Hispanic patient population, which is completely different from South Africa.

The Massachusetts Institute of Technology’s (MIT) Laboratory for Computational Physiology in collaboration with Beth Israel Deaconess Medical Centre developed MIMIC-IV. It contains anonymous records from patients admitted to hospitals and ICUs within 24 hours of visiting an Emergency Room (ER) in the USA between 2008 and 2019.

The AI model in the study used MIMIC-IV’s detailed information about patient age and gender, admission characteristics, vital signs and previous psychiatric diagnoses.  These are routinely recorded in hospital systems and helped the model identify patterns associated with SRB.

Tuning the model, outputs for clinicians

The study’s model did not include job loss, bereavement, drug, or alcohol-related observations from medical staff, but can be included in future versions, says Mokheleli.

Future models will use bigger data sets, potentially different AI algorithms, and adjusted techniques to improve the performance of the model, adds Dr Tebogo Makaba, a study co-author and Mokheleli’s PhD co-supervisor. Dr Makaba is  Senior Lecturer in the Department of Applied Information Systems at the UJ CBE.

Future versions of the model will also require the ‘patient risk threshold’ to be adjusted for each hospital, says Mokheleli, since patient populations vary.

In addition, the AI model’s current output - confusion matrices and SHAP (SHapley Additive exPlanations) plots - is hard to understand.

Says Mokheleli: "A SHAP plot shows how a model made a decision about one patient. We are busy designing a user interface that clinicians can interact with and get answers they can understand.”

###

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The authors acknowledge the PhysioNet team for providing access to the MIMIC-IV database and supporting reproducible research in critical care and emergency medicine.

 

Yeast and electricity convert CO₂ into valuable chemicals



Researchers develop a new process that captures CO₂ and produces a valuable bio-based chemical




Helmholtz Centre for Environmental Research - UFZ

Graphic

image: 

This study demonstrates a novel hybrid system that integrates electrochemical CO2 reduction with microbial biosynthesis of a valuable chemical in a single reactor.

view more 

Credit: Izadi et al. Angew. Chemie / Graphic: Simone Bachleitner, Paniz Izadi






Itaconic acid is a bio-based organic acid used to manufacture biodegradable plastics, textiles, adhesives, and resins. It is considered a promising sustainable alternative to petroleum-based chemicals. Worldwide, around 70,000 t of itaconic acid are produced annually, and production is increasing. “On an industrial scale, itaconic acid is produced biotechnologically from sugar using a fungus. We have now developed a process that uses CO₂ instead of sugar as a starting material so that we can put emissions to productive use”, explains Prof. Falk Harnisch, an electrobiotechnologist at the UFZ and last author of the study.

In collaboration with researchers from BOKU University in Vienna, the UFZ team has developed a method that combines electrochemical and microbial processes in a single bioreactor. But how exactly does this electrobioreactor work? And how did the researchers come up with the idea of combining the use of CO₂ with the production of itaconic acid in the first place?

A previous study by BOKU University showed that itaconic acid can be produced biotechnologically from methanol, a type of alcohol. Because microorganisms can metabolise a wide range of substances, the team led by Harnisch tested a chemically similar substance that can be produced relatively easily from CO₂. “Electricity, ideally from renewable sources, can be used to convert CO₂ into formate, which is chemically related to methanol because both contain just one carbon atom. We therefore wanted to investigate whether the yeast could also use formate as an energy source for itaconic acid production”, says Dr Paniz Izadi, a researcher in Harnisch’s UFZ team and first author of the study.

The design of the electrobioreactor in which electrochemical and microbial processes take place simultaneously presented a number of challenges. The electrochemical conversion of CO₂ into formate works best at elevated temperatures and an alkaline pH in the presence of as few other chemical substances as possible. The genetically engineered yeast strain Komagataella phaffii, which is used in the bioreactor, requires entirely different conditions: temperatures of around 30°C, a neutral pH, and a nutrient-rich medium. “In order to ensure that both the electrochemical and microbial processes took place and operated as efficiently as possible, we had to find a compromise between their different requirements in the electrobioreactor”, explains Dr Simone Bachleitner, Senior Scientist at the Institute of Microbiology and Microbial Biotechnology at BOKU University.

The experiment was successful, and the yeast produced itaconic acid from CO₂ using electrochemically produced formate as an energy source. “We successfully combined electrochemical and microbial processes in our bioreactor. A self-contained process chain offers major practical advantages because it eliminates the need to transport intermediate products between separate facilities. Although the yield of itaconic acid was lower than when sugar or methanol was used, our study clearly demonstrates that CO₂ can be converted into complex and valuable biochemicals”, says Harnisch.

The researchers hope that their new process will provide a model for practical applications and contribute to the development of industrial-scale facilities. The researchers now plan to optimise the processes in the electrobioreactor and investigate their suitability for other microorganisms and bio-based products.