The hip bone is one of the most reliable indicators of sex in the human body due to the fact it is the most dimorphic bone. Probabilistic Sex Diagnosis (developed by Murailet al., 2005) is a method based on a worldwide hip bone metrical database and relies on the actual physical bone for analysis. Sex is determined by comparing specific measurements taken from each specimen using sliding calipers and computing the probability of the specimens being female or male. In forensic science it is sometimes not possible to sex a body due to corpse decay or injury. Skeletinization and dissection of a body is a laborious process and desecrates the body. The current study aimed to see if it was possible to virtually utilise the DSP method to avoid this process. Forty-nine innominate bones of unknown sex were obtained from ULB. Bones were analysed by two researchers using the manual DSP method and a good correlation was found between researchers. CT scans of available bones were analysed to obtain three-dimensional (3D) virtual models using a commercially available software (Amira, www.amiravis.com). Available models were imported into a customized software programme called lhpFusionBox (developed at ULB from the MAF open-source library). lhpFusionBox is an advanced musculo-skeletal software which includes many operations relevant to Biomechanics. It also enables distances to be measured via virtually- palpated bony landmarks. DSP measurements were then obtained from the located bony landmarks. There was 100% accuracy between the manual and virtual DSP analysis. To further test the method 30 virtual bones of known sex were analysed (researchers had no prior knowledge of sex before analysis). There was found to be a 97% accuracy rate with only one bone leading to a wrong determination. These

International Redaction Board, Abstract of an Oral Presentation or a Poster