The radiologic assessment of lung infiltrates by conventional anterior-posterior chest radiographs has been embedded in the concept of acute respiratory distress syndrome (ARDS) since its original description by Ashbaugh and colleagues [1]. In fact, the presence of bilateral lung infiltrates as detected by chest radiograph, in association with ''refractory'' hypoxemia and impaired mechanics of the respiratory system (stiff lung), was one of the mandatory criteria for ARDS diagnosis. In the mid-1980s, however, the first reports describing ARDS by using CT scanning changed the view of the syndrome [2][3][4]: the ARDS lung appeared to be affected by the disease process nonhomogeneously, with the CT densities distributed mainly in the dependent lung regions. The quantitative analysis of CT scanning gave new insights concerning the pathophysiology of the syndrome and provided a firm rationale for modifying the mechanical ventilation in use at that time [5][6][7]. This article discusses in some detail what the authors believe is or should be the role of the CT scanning in the diagnosis and therapy of ARDS.
Quantitative analysis of CT scanningBecause the authors believe that the quantitative analysis of CT scanning is of paramount importance for ARDS diagnosis and for a rational setting of mechanical ventilation, a brief description of the technique and its limits is necessary. (A more detailed description of the quantitative analysis may be found elsewhere [7,8]).
Physical principles of CT scanningThe digital image produced by the CT scan is based on the measure of the attenuation coefficient (m) (ie, the reduction of the radiation intensity upon passage through matter). Traversing from one side of the patient to the other, the X-ray beam is attenuated by all the voxels (volume elements of the tissue) through which passes. The intensity of the emerging X-ray is described by the Lambert law of absorption, that is,where I is the intensity of the emerging X-ray beam, and I 0 is the intensity of the X-ray beam at the source. By measuring the intensity I and I 0 , the CT calculates the integral over the function m of the attenuation coefficient along the X-ray beam. Then, through a different mathematical algorithm (usually filtered back projection), a given attenuation number m is assigned to each voxel (m voxel ). This attenuation number primarily represents the density of the voxel (ie, the ratio of mass to the volume) and is expressed as the CT number [9], which relates to the density of the water (m water ):CT ¼ 1000 Â m voxel À m water m waterConventionally a CT number equal to zero Hounsfield units (HU) indicates that density equals