A Bayesian approach to noise removal in Dynamic MRI of Lung

Introduction We have previously shown that dvmunic multislice MRI sequences acquired contmuously before, during, and after injection of contrast agent give information that may be useful in diagnosing lung diseases like pulmonary emboli (PE), chronic obstructive pulmonary disease (COPD), and pneumonia [1,2]. Another potential pulmonary MRI technique is ventilation imaging based on breathing interleaved 100% oxygen and room air [3,4]. We have previously presented a novel method for automatic correction of the deformation of the lung, as a first step towards pixel-by-pixel analysis [5]. However, the low signal-to-noise ratio (SNR) in the lung is another obstacle to overcome. We therefore propose a novel, Bayesian approach to noise reduction.