Techniques in computer-aided diagnosis and their application in clinical investigation of bronchial systems

Respiratory diseases constitute a major preoccupation for the medical community, due to their worldwide extent, their high incidence in the industrialized countries and an important mortality rate. In this context, early diagnosis is the key issue for the patient healthcare policy. After its introduction in clinical routine during the last decade, helical computed tomography (CT) became rapidly the recomended imaging technique for assessing airway disease. With the advent of multidetector row CT (MDCT), high resolution images of the airways were possible to be acquired throughout the whole thorax, in a single breath hold. But the advantage to benefit of high-quality data for pulmonary investigation was counterbalanced by the large amount of data the clinician had to deal with. Computer-aided diagnosis (CAD) techniques are now proposed in routine investigation. This paper aims at presenting an overview of the advances in the CAD techniques designed for bronchial systems analysis. According to their investigation ability, both basic and advanced methods are addressed throughout this presentation. From direct visualization, to complex segmentation, interaction, navigation, simulation and quantification issues, the challenges raised by the airway pathology investigation are discussed and various solutions are presented and illustrated.