In the latest issue ofPhysiological Measurement, a group of EIT experts has published a review of lung EIT – its current status and future prospects – based upon presentations at recent biomedical EIT conferences. The authors note that lung EIT is at a critical transition point, with commercial devices recently introduced to the market and growing clinical interest. As such, they emphasize the need to move from validation studies to prospective studies that use EIT to diagnose and treat patients (Physiol. Meas. 33679).
Lead author Andy Adler, from Carleton University in Ottawa, Canada, explained the incentives for writing the review. "At the 2009–2011 biomedical EIT conferences, the 'special lung session' speakers presented some very good talks, and I wanted to try to collect their thoughts," he said. "The other main motivation was the high level of excitement that we're now seeing for lung EIT in the critical care medicine community. I was a little worried that the medical and engineering communities were not understanding each other well. The review paper was aimed to help bridge this gap."
Where are we now?
Adler and co-authors begin by taking a look at current lung EIT applications. Proposed fields of use fall into two categories: ventilation monitoring, and monitoring of perfusion and gas exchange. The former – which includes measurements of tidal volume distribution, intrathoracic gas volume and respiratory system mechanics – is fairly advanced, with some confidence in its accuracy and repeatability. The latter application is less mature, but also shows promise.
EIT instrumentation has now shifted from research-driven development to the release of commercial thoracic EIT systems. All of these use a small number of electrodes placed in a transverse plane around the thorax, with current driven across electrode pairs as differential voltages are measured. Image reconstruction approaches are based on linear, one-step reconstruction algorithms.
The authors also detail some promising EIT hardware advances that have been proposed but not yet systematically validated. Examples include 3D EIT systems, which use multiple bands of electrodes to create a 3D image of conductivity changes in a larger section of the lungs, and multi-frequency systems that use several stimulation frequencies to detect and image characteristic differences between tissue types.
On the image reconstruction side, algorithms are being developed that preserve edges and use robust data norms to create higher quality images with reduced sensitivity to data errors. In particular, the authors highlight novel image analysis algorithms that emphasize the analysis of dynamic behaviour.
Where do we want to go?
The next big challenge, says Adler, is to move from looking at what EIT can measure, to determining how it can be used to guide treatment. In the longer term, he envisions that EIT could become part of standard operating procedures, in which EIT-derived parameters (percentage of lung collapse, for example) are monitored and ventilation settings modified accordingly. EIT values could also be used within automated systems to control ventilation.
The authors identify a set of requirements needed to enable clinical EIT use. These include: availability of EIT devices at reasonable cost; standardization of data and image formats; robustness against electrode contact problems and electrical interference; an intuitive software interface; and standardized EIT protocols.
They also collate a list of feasible clinical EIT applications. In one example, EIT is used to help select ventilator parameters that provide adequate gas exchange while reducing ventilator-induced lung injury. EIT can be used to generate early alarms if the lung's condition deviates from a previous optimum, or following the onset of dangerous conditions. EIT could serve as a teaching tool by making aspects of lung behaviour visually clear and intuitive. Finally, EIT could manage the weaning of patients off ventilatory support and enable therapies that are otherwise unachievable.
How do we get there?
So what's needed to bring EIT to clinical use? The authors recommend a programme of clinical research and engineering development centred on applied problems. Research must focus on relevant outcomes: reducing the time needed to achieve treatment goals such as improved oxygenation index, for example, or minimizing adverse events such as pneumothorax and ventilator-induced lung injury.
It's also important that clinical studies look into EIT-measured variables for guiding therapy. For example, an EIT measure that a patient's lungs are 30% collapsed may provide insight into which therapeutic intervention to employ. Future work should also include animal experiments using EIT-guided therapy to treat models of lung injury.
In terms of hardware requirements, the most pressing need is for equipment to be available in sufficient quantities at reasonable cost. It's also necessary to understand what electrode types and contact fluids work best with which patients, and improve patient interfaces to allow quick and reliable electrode placement, as well as automatic detection and correction of any set-up errors.
EIT images should be analysed using sophisticated signal processing algorithms to generate new parameters and exploit temporal values. There's a need to validate reconstruction algorithms and develop ways to automatically select regions-of-interest in lung images. Clinical EIT systems also require user-friendly software interfaces, and a means of representing EIT images and data in a clinically useful manner, using a standard format.
To solve all these challenges, the authors emphasize the need for improved collaboration between clinical and engineering teams. "I think we're at a bit of a tipping point," Adler told medicalphysicsweb. "If we don't manage to make EIT reliable, it will be seen as 'yet another promising technology that's not ready for the clinic'. On the other hand, EIT has the potential to provide insight into what's going on in the lungs of difficult patients. In five years, we will hear doctors say: 'this is a complicated case... make sure you put the EIT belt on'."
• This review article is part of a focus issue inPhysiological Measurement covering the 12th International Conference on Biomedical Applications of Electrical Impedance Tomography. The conference was held in the UK and hosted at the University of Bath.