Quantitative Approach to Assessing Tracheal Stenosis in Rabbits Based on an Open-Source Image Processing Software
摘要
Occlusion area and location are relevant information for tracheal stenosis diagnosis and treatment. The existing classification systems are, usually, qualitative or offer only approximations of the aforementioned information. The present work applies two quantitative methods capable of pointing to stenosis presence and determining its occlusion area and location: measuring the area occupied by the segmentation of the airway (slice area) as a function of axial coordinate, and segmentation volume. The methodology was validated by comparing the trachea of six rabbits, divided into a control and an intervention group, at two different moments: before and after inducing a lesion. Slice area was evaluated at the expected stenosis position and at the point of maximum occlusion, while volume was used to evaluate the airway geometry globally. The variation of slice area and volume before and after the lesion were used as stenosis assessment metrics. The greatest difference of slice area variation between both groups at the expected stenosis position was 67.22%. For the volume, the greatest difference observed was 27.06%. The proposed methodologies were able to point to stenosis presence, where slice area was capable to give a better quantitative assessment of the airway narrowing, while volume was an effective stenosis indicator.