Electronics from a car airbag system is now being used to remotely monitor the activity of patients, writes Anna Nolan.
The electronics used to make car aribag systems pop in the event of an accident is now also being used as a way to monitor the the activity of patients recovering from stroke and other illnesses that affect mobility.
Airbag accelerometers form a key part of an easy-to-wear monitoring system developed by a group of researchers and medical specialists from Limerick hospitals and the University of Limerick (UL). The group uses the angle-measuring ability of accelerometers to determine the patient's angle relative to the ground at any given time. Regular physical activity, while important for everyone, is particularly important for post-stroke patients and people regaining mobility after falls or accidents, according to the research team. This makes it very important to know a patient's pattern of movement.
The team, including Drs Karen Culhane and Gerard Lyons from UL, Dr Diarmuid Hilton and Prof Declan Lyons from the Rehabilitation Centre at St Camillus' Hospital, and Prof Pierce Grace, from the Vascular Surgery Department in the Mid-Western Regional Hospital, tested the system both in hospital and in home environments.
The monitoring system consisted of two airbag accelerometers made by Analog Devices, a data storage device that could be worn while walking around or resting, and the related cables to connect everything.
"These accelerometers only cost about €10 to €20 each, so we are leveraging on that technology," says Lyons, director of the biomedical electronics laboratory at UL.
One accelerometer was placed on the patient's trunk and another on one thigh, each providing data on angles. Using custom-designed programs, a comparison of the two angles gave body position, and a summary of the minute-by-minute activity of each patient was produced.
Five residents in the rehabilitation clinic aged between 52 and 90 were tested for four days, with a researcher shadowing them and logging their movements. The data from the accelerometers was analysed using a program that allowed trunk and thigh angles to be set in order to distinguish between sitting, standing, lying and moving.
The threshold angle at which the software decided between two options - whether the person was lying or sitting, for example - proved crucial. Previous researchers had used an angle of 45 degrees, which gave a true position accuracy rate of only 75 per cent on average when compared with the observer's record of activity.
"The problem was that people who are unwell tend to slump, and so could be leaning more than 45 degrees forward, although actually sitting," says Lyons.
The team then used a "best estimate" for the threshold angle, and this improved the accuracy to an average 93 per cent. They also raised the accuracy of detecting the difference between static activity (sitting, standing and lying) and dynamic activity to an average of at least 92 per cent.
They went on to test one patient in the home, also with human shadowing, and the technique proved just as successful in that setting.
The ultimate aim is to produce a simple system for remotely monitoring an elderly patient's pattern of movement.
Another Biomedical Electronics Laboratory researcher, Cliodhna Ní Scanaill, is looking at adding mobile telephony to the existing prototypes. She is at the Medical School at Grenoble for the summer, testing the system with a group of researchers working on technology associated with the "smart home" concept.