Researchers at Queensland University of Technology have developed a wearable sensor system that detects early signs of worsening heart failure days before symptoms become severe enough for patients to seek medical attention. The device monitors multiple physiological signals and uses machine learning to identify patterns indicating deterioration.

Heart failure affects more than 500,000 Australians and leads to frequent hospitalisations when the condition worsens. Catching deterioration early allows medication adjustments or other interventions that prevent hospital admission. But patients often can’t identify subtle changes until symptoms are severe.

The QUT device addresses this by continuously monitoring heart sounds, respiratory rate, chest impedance indicating fluid accumulation, and activity levels. An algorithm trained on data from hundreds of heart failure patients identifies combinations of changes that reliably predict worsening within three to seven days.

Clinical Problem

Heart failure is a chronic condition where the heart can’t pump blood effectively enough to meet the body’s needs. It’s manageable with medication, but the condition typically worsens over time with periodic acute exacerbations requiring hospitalisation.

These exacerbations are expensive, accounting for the majority of heart failure treatment costs. They’re also dangerous, with each hospitalisation increasing risk of subsequent events and death. Preventing hospitalisations by catching deterioration early improves outcomes and reduces costs.

Current monitoring approaches rely on patients weighing themselves daily and reporting significant weight gain, which indicates fluid retention. But weight changes don’t always correlate tightly with worsening heart failure, and many patients don’t comply consistently with daily weighing protocols.

Some patients use implanted monitors that detect fluid accumulation in the lungs more directly. These work well but require minor surgery for implantation and cost several thousand dollars per patient. A non-invasive wearable alternative would be suitable for more patients.

How the Device Works

The QUT sensor is about the size of a small smartphone and adheres to the chest using medical-grade adhesive. It contains accelerometers to measure heart sounds and respiratory motion, electrodes to measure chest impedance, and a processor running analysis algorithms.

Battery life lasts approximately five days per charge. Patients remove the device to shower and recharge it. Data uploads to cloud servers via smartphone app, where more sophisticated analysis algorithms run and alert clinicians if concerning patterns emerge.

The heart sound analysis detects subtle changes in the timing and intensity of heart valve closures that indicate worsening function. Respiratory patterns provide information about breathing effort and abnormalities. Chest impedance correlates with fluid accumulation in the lungs, one of the key signs of worsening heart failure.

None of these measurements individually provides reliable early warning. But the combination, analysed through machine learning algorithms, achieves much better predictive accuracy. The algorithms learned to identify patterns characteristic of deterioration by training on data from 300 heart failure patients monitored for an average of eight months.

Clinical Trial Results

A trial involving 120 heart failure patients compared the device to standard care over six months. Half the patients used the sensor continuously, while the control group received usual monitoring including regular clinic visits and daily weighing.

The sensor group had 47% fewer heart failure hospitalisations than controls. When deterioration was detected, clinicians typically adjusted medications, sometimes increased monitoring frequency, and occasionally scheduled urgent clinic visits. These interventions prevented many hospitalisations.

False alarm rates were acceptably low. The system flagged potential deterioration approximately once every three months per patient, and about two-thirds of those alerts corresponded to genuine clinical deterioration. The remaining third were either false alarms or subtle changes that resolved without intervention.

Patients generally accepted wearing the device, though some found the adhesive irritating and others disliked the charging routine. Compliance rates averaged 85%, meaning patients wore the device as directed 85% of the time. That’s sufficient for clinical benefit but suggests room for user experience improvements.

Dr Michael Chen, the cardiologist who led the trial at the Prince Charles Hospital in Brisbane, said the results support broader deployment but noted that the technology needs integration with existing clinical workflows. Alerts must route to the right clinical staff and trigger appropriate response protocols.

Commercialisation Path

QUT has licensed the technology to Melbourne-based medical device startup Cardio-Sense, which is raising capital for regulatory approvals and manufacturing scale-up. The device requires Therapeutic Goods Administration approval as a medical device before Australian commercial deployment.

TGA approval for this class of device typically takes 12-18 months, depending on how complete the regulatory submission is and whether the agency requests additional data. The QUT trial data should satisfy efficacy requirements, but manufacturing quality systems and risk analyses need detailed documentation.

International regulatory approvals will follow Australian approval. The United States FDA and European CE marking are priorities for market expansion. US approval often takes longer than Australian but opens a much larger market.

Pricing strategy remains undetermined. The device must be affordable for health systems that will pay for it but expensive enough to support the company’s business model. Some medical devices are sold directly to healthcare providers, while others are reimbursed through payment systems like Medicare. Each approach has implications for pricing and market development.

Market Context

Several companies worldwide are developing remote monitoring solutions for heart failure. American Well and Livongo offer virtual care programs with periodic monitoring. Zoll Medical makes wearable defibrillators with some monitoring capabilities. Cordio Medical is developing a voice-analysis system that detects deterioration.

The QUT device differentiates itself through continuous multi-parameter monitoring and the machine learning analysis combining different signals. Whether that proves decisive commercially depends on comparative effectiveness studies and how healthcare payers value the technology.

There’s also the question of how remote monitoring affects clinical workflows. Cardiologists and heart failure nurses already manage large patient panels. Adding continuous monitoring data could improve care but also increases workload. Implementation needs to account for workflow integration.

For healthcare providers evaluating remote monitoring technologies, assessing not just clinical effectiveness but also implementation requirements matters. Technologies that perform well in trials sometimes struggle in routine practice because workflow integration proves difficult.

Future Directions

Cardio-Sense plans to develop a next-generation device that’s smaller and requires less frequent charging. Miniaturisation and improved battery technology should enable week-long battery life in a device half the current size.

There’s also interest in expanding monitoring to other chronic conditions. Chronic obstructive pulmonary disease and kidney failure could potentially benefit from similar continuous monitoring approaches. The company is conducting feasibility studies to evaluate those opportunities.

Integration with other digital health tools represents another development direction. Combining continuous monitoring data with electronic health records, medication management systems, and telehealth platforms could create more comprehensive chronic disease management.

Australian researchers continue investigating improved analysis algorithms that might detect deterioration even earlier or reduce false alarm rates. Research groups at University of Melbourne and University of Sydney are collaborating with Cardio-Sense on algorithm development.

For the 500,000 Australians living with heart failure, technologies like the QUT sensor offer the potential for better outcomes and fewer hospitalisations. The path from research prototype to widely deployed clinical tool takes years and requires navigating regulatory, commercial, and clinical adoption challenges. But the clinical trial results suggest that path is worth pursuing.