A Study for the Determination of Sex by Multidetector Computed Tomography of Sternum using Discriminant Function and Logistic Regression

Abstract

Post-mortem investigations of skeletal remains as well as radiographs from living individuals provide useful information for the discrimination of sex. Our study aimed to find out a mathematical model to differentiate gender based on greater degree of accuracy than the anthropological measures taken from the sternum obtained from cadaver dissection. The study was performed on 108 adults who were brought for examination of chest due to various medical reasons. Their age ranged between 18 and 80 years.The cases were selected randomly after considering the inclusion and exclusion criteria. Sternal measurements were taken by studying CT (Computed Tomography) scans. Of these cases, 73 were males and 35 were females. The discriminant function equation (Df) = 0.071 Manubrial Length +0.075 Manubrio-Sternal Length +0.036 Width at S1 +0.037 Width at S3-11.367 (Constant). Overall 80.6% of the sample was correctly classified into their group. This study revealed that measurements from CT scan of sternum can be used to differentiate between sex of individuals which adds to a great advantage in forensic anthropology.