AI experts consider ethical use of video technology to support patients at risk of falls

Video-enabled glasses have the potential to support patients at risk of falls by allowing medical staff to monitor how patients move around their homes and communities. But with privacy concerns at the forefront of this new technology, academics at Northumbria University have published the latest information on the ethical use of AI to ensure video footage is hidden to ensure patient privacy. We conducted cutting-edge research.

Traditionally, patients at risk for falls have been assessed based on information provided by the patients themselves, such as in diaries or during short in-hospital observation appointments. However, these do not provide clinical teams with objective digital data about how patients move around outdoors and in their home environments, where falls are most likely.

Attempts at waist-worn inertial wearable technology, similar to those used in smartwatches, to track a patient's walking motion, known as gait, have proven beneficial, but it is important to understand where the patient is. There are still pitfalls due to lack of contextual information about who is walking with them and what activities they are engaging in.

Experts are investigating how to improve and personalize patient care for patients at risk of falls due to illness or aging, by improving assessments and possible causes of abnormal gait data collected. We have been eagerly searching for a way to understand in more detail whether or not this is the case. by wearable devices.

Asking patients to also wear video-enabled glasses will help you understand how they move in relation to their surroundings, including obstacles and other hazards, where they are at a given time, and what puts them at risk of falling. Get more accurate information about what you are exposed to.

However, while the use of video technology has many potential benefits, patients and their families wearing these video-enabled glasses must be able to maintain privacy.

To test how these privacy concerns can be overcome, a group of computing and digital health experts analyzed personal data and information captured with video glasses, such as photos and footage from around the house. We ran a technology pilot to test the application of a new AI software that can blur. Children and confidential documents.

They discovered that AI software can analyze raw video footage and detect and blur details such as faces, text, and laptop or phone screens to ensure patient privacy.

Their research is currently npj digital medicine.

Dr Alan Godfrey, Associate Professor of Computer and Information Science at Northumbria University, said: 'As you can imagine, there is huge variation in the way people move when completing different tasks.

“The data and information provided by an inertial wearable device worn while someone is wandering around the house is faster than when the same person is outdoors, trying to get somewhere, or keeping up with someone. It's necessarily different when you're walking, which means using an inertial wearable alone helps, but it doesn't tell you anything about the situations in which a fall might occur.

“It's important to be completely clear about what the environment and people are doing.”

He added: “We would like to assess how new developments in AI could enable the provision of video-enabled glasses that allow medical staff to observe patient movements in real-world environments over extended periods of time without compromising privacy.” I thought about it,” he added.

This paper combines information from wearable devices that record gait data with footage captured by video glasses that is optionally obscured through the ethical use of AI to help clinicians see patients around them. It has been demonstrated that it is possible to gain a more comprehensive understanding of how the system works. .

This could lead to significant improvements in the accuracy of patient fall risk assessment and the decision-making process regarding patient care.

Regarding this study, the principal investigator and Ph.D. Jason Moore, a student in Northumbria University's School of Computer and Information Sciences, said: “Traditionally, using video in the home has been limited to patient populations due to privacy concerns as other things can be seen on camera. “It has caused anxiety,” he said.

“However, by using AI software that can identify and hide personal and sensitive information, we can effectively capture contextual information that can provide a deeper understanding of anomalous gait data, while also enabling the use of video technology in medical settings. Overcome concerns patients have about their home.

“The benefit of providing this contextual information is that it allows clinicians to see a complete picture of each individual patient, ultimately allowing them to provide a more informed plan of care and better It has the potential to keep many patients at home for longer.”

The research involved experts from Northumbria's School of Computer and Information Sciences. Nursing, midwifery and health. Representatives from Sport, Exercise and Rehabilitation, Northumbria Healthcare NHS Foundation Trust and Cumbria, Northumberland, Tyne & Wear NHS Trust will also be in attendance.

Dr Godfrey continued: “The proposed application of this technology is unique; the way it works means that the raw footage is never seen by the clinician. Gait asymmetry, or asymmetry in the way a patient walks or moves, is a sign of increased fall risk and can be invaluable. is.

“Without this context, patients with many clinical conditions, for example those with Parkinson's disease, those who have had a stroke, or even those suffering from frailty, would be classified as being at high risk of falling and would eventually be displaced. Replacing this with a community-based approach also reduces the pressure on patients to attend bespoke clinics for monitoring. ”

“This could go a long way in helping us truly understand a patient's fall risk and provide personalized care solutions that best suit the individual and their needs.”

Following this first research paper, the team plans to investigate the risk of habitual falls in people with Parkinson's disease, and are currently recruiting people with Parkinson's disease to wear the technology at home or in the community. This will allow the research team to refine and improve their AI algorithms, which in the future will help inform personalized approaches to reducing the risk of falls in people with Parkinson's disease.

Speaking about the project, co-author Professor Richard Walker, consultant physician at Northumbria Healthcare NHS Foundation Trust, said: 'People with Parkinson's disease have increasing mobility problems as their disease progresses. They can lead to major injuries such as broken bones, so anything we can do to prevent them will be of great benefit.

“It is hoped that this new technology will provide unique information about fall situations and allow us to advise on the most appropriate preventive measures.”