| GREIT: towards a consensus EIT algorithm for lung images, (2008) | |||||||||||
Abstract | |||||||||||
| Recently, electrical impedance tomogra- phy (EIT) has begun to see a signi¯cant clinical in- terest for monitoring of ventilated patients. The key capability of EIT is to provide real-time images of the distribution of ventilation in the patient's lungs. However, most clinical and physiological research in lung EIT is done using older and proprietary algo- rithms; this is an obstacle to interpretation of EIT results because the reconstructed images are not well characterized. To address this issue, we are devel- oping a consensus linear reconstruction algorithm for lung EIT, called GREIT (Graz consensus Reconstruc- tion algorithm for EIT). This algorithm is being de- veloped in three phases: 1) selection of the "ingre- dients" and evaluation methodology (this paper), 2) evaluation and experience with GREIT variants, and 3) consensus and definition of the GREIT algorithm. Algorithms evaluation criteria are identified to be: a) quantitative output for all positions, b) reconstructed position error (low and uniform), c) resolution (small PSF, uniform, few artefacts), d) good noise perfor- mance, e) low sensitivity to electrode and boundary movement, f) good performance on clinical and exper- imental data. This approach represents the consensus of a large and representative group of experts in EIT algorithm and clinical applications. All software and data to implement and test GREIT will be made avail- able under an open source license which allows free research and commercial use. | |||||||||||
Publication details | |||||||||||
| |||||||||||