HemoVision is software package that accelerates the process of bloodstain impact pattern analysis. To assess its accuracy, we set up a validation study in which we compared HemoVision's Area of Origin (AO) estimates to ground truth values. Moreover, we compared HemoVision's accuracy to that of HemoSpat and manual analyses performed by a trained bloodstain analyst.

The tangent method is often used to estimate the AO of impact patterns, although it has been shown to provide biased results because it uses front-view projections of bloodstain trajectories to first determine the Area of Convergence (AC). HemoVision therefore uses a novel method that works directly in three dimensions, eliminating the need for front-view (or top-view) projections. This method is formulated as an optimisation problem that minimises the total sum of squared angles between the estimated and the corrected linear trajectories, the latter of which are defined as the lines connecting the AO center with the stain coordinates.

To assess the accuracy of this novel AO estimation method as well as HemoVision's overall accuracy, we created 10 impact patterns with known AO coordinates. These were then compared to HemoVision's estimated AO coordinates. HemoVision's accuracy was also compared to HemoSpat and manual analyses performed by a trained bloodstain analyst. A summary of the results are given in the figure below.

For more information about the proposed AO estimation method or results from the validation study, please refer to the original research article here.

Based on: Joris, Philip, et al. "Bloodstain impact pattern Area of Origin estimation using least-squares angles: A HemoVision validation study." Forensic Science International 333 (2022)