Study of Anatomical Variations of Para nasal sinus (PNS) using Multislice Computed Tomography (CT) in Rhinosinusitis

The study aimed to characterize anatomical variations of Para nasal sinus (PNS) in Rhino sinusitis depending on the classification processes of CT Brain using Interactive Data Language (IDL) program as platform for the generated codes. This retrospective study included 392 patients, 272 males, 120 females, Their ages ranged from 18 to 68 years, the majority of cases were found in age group from 28 to less than 38 years 154 Patients (39.3 %), referred for CT brain at Riyadh Hospitals, SaudiArabia, mostly from Consultant Radiologists Diagnostic Center selected randomly. The results of the classification showed that incidence of PNS diseases was more in males (69.4 %) as compared to females (30.6 %). The predominant clinical presentations were headache 276 Patients (70.4 %), and the most the sinuses areas were classified well from the rest of the tissues although it has characteristics mostly similar to surrounding tissues. Several texture features are introduced from the first order statistics and the classifications score matrix was generated by linear discriminate analysis and overall classifications accuracy of sinuses regions found was 94.3 %, and the classification accuracy of normal sinuses 100 %, the classification accuracy of abnormal sinuses measured as 89.7 %, and for spine as 86.9 %, While the bone showed a classification accuracy of 100 %. These relationships are stored in a Texture index that can be later used to automatically annotate new CT with the appropriate characterized anatomical variations of PNS, which will help defining Rhino-sinusitis.