The workload in radiology departments has been increasing substantially over the last few decades. This is due to the greater need of tomographic examinations, as well as the increasing number of slices in each examination, determined by the advancements in tomographic technology. In order to ameliorate this, it is necessary to implement means of optimising the workflow of the diagnostic radiologist. Among them the most widely spread and easily accessible are special methods for visualization and image postprocessing – multiplanar reformats, maximum intensity projections, volume rendering techniques, and segmentation. They enable easier differentiation of unclear findings, faster and more reliable discovery of fine small calibre lesions and thrombi, improved spatial orientation and pre-operative planning, as well as acquisition of reproducible and reliable medical scientific measurements. These methods are available as builtin modules in most medical imaging software packages (including ones with an open source) and are an integral part of radiological interpretation, saving time and effort. In the future they can be reinforced with highly specialized artificial intelligence, which could make automatic measurements and locate a specific type of finding.
Read full abstract