Philip Joris - Profile on Academia.edu (original) (raw)
Papers by Philip Joris
Forensic Science International, 2015
1.2. A novel approach to automated bloodstain pattern analysis using an active bloodstain shape model
Journal of Forensic Radiology and Imaging, 2014
ABSTRACT Objectives Conventional blood pattern analysis is a tedious and time- consuming process ... more ABSTRACT Objectives Conventional blood pattern analysis is a tedious and time- consuming process due to the many actions that must be performed by the pattern analyst. Digital photographs can now be employed to aid in the process of scene analysis, by letting computer programs perform some of the required steps. However, current computerized methods are all based upon the assumption that any stain can be approximated by a simple ellipse, while manual work is still required. Materials and methods This work presents a novel approach, employing a regressed active shape model to approximate bloodstains. This Active Bloodstain Shape Model (ABSM) uses a regression to correlate the shape of a stain׳s approximation to its impact angle. This model is then deployed in a software pipeline, aiming to eliminate any user-input from the process. Fiducial markers are positioned in a crime scene, allowing for fully automated pattern analysis. Images are cleaned of any perspective distortions, after which stains are segmented and analyzed using said model. A robust principal component analysis (PCA) is then used to analyze the intersections of the reconstructed flight paths. Results Experimental results have demonstrated that the ABSM is able to better approximate stains, while accuracy on predicted impact angles is increased. Both simulated as well as real crimes scenes have confirmed the working of the automated pipeline. Conclusion Because the ABSM learns from experimental data, it is able to better approximate stains, resulting in an increased accuracy. The addition of the software pipeline removed the need for almost all user-input.
3.5. The feasibility of measuring body weight on CT images and the first steps in anatomical mirroring
Journal of Forensic Radiology and Imaging, 2014
ABSTRACT Objective To estimate body weight using the data of a post mortem total body CT scan and... more ABSTRACT Objective To estimate body weight using the data of a post mortem total body CT scan and establish its efficacy in a forensic context. Material and method PMCT data were used to compute fat, soft tissue and bone volume using in house developed software based on density and HU. The digitally calculated body weight was then compared with the measured body weight on autopsy. Results Analysis of the obtained results showed a good correlation between the measured body weight and the CT-derived body weight. The highest discrepancies were noticed in children and decomposed bodies presumably due to different bone mineralization in children versus adults and putrefactive gas formation and liquefaction in decomposed bodies. In general, there was no distinct consistent over- or underestimation. In case of partial scans, where a body part was not scanned or was missing, mirroring the unaffected side was seen to give equally good results. Conclusion The weight of a body can be sufficiently estimated using volume measurements of different body tissues obtained from the PMCT data. This technique can be applied when concern is raised about the documented body weight, or if weighing was omitted all together. Furthermore the aforementioned body weight estimates can be useful in case of mass casualties, for victim identification purposes, or in case of traumatic deaths (e.g. dismemberments).
Journal of Forensic Radiology and Imaging, 2014
Objectives: Traditional Bloodstain Pattern Analysis (BPA) estimates impact angles for bloodstain ... more Objectives: Traditional Bloodstain Pattern Analysis (BPA) estimates impact angles for bloodstain spatters by employing ellipses and an inverse sine. This approach is based upon a simplification of the stain formation process, and ignores the physical properties of blood, and its intricate interactions with the target surface. This research presents a data-driven approach, starting from experimental bloodstain spatters. No assumptions about the shape of stains or about the relation between the shape and their impact angle are made a priori. Materials and methods: Experimental bloodstain data is gathered, after which shape variations are extracted by employing an Active Shape Model. A non-linear regression is then used to explain the relation between stain shape and impact angle. Results: Experimental results show that traditional width-length ratios may deviate from the assumed inverse sine. The proposed approach, on the other hand, closely follows the experimental data through its regression. Additional experiments have shown an increased accuracy on impact angle estimates for the proposed approach when compared to automatic ellipse fitting. Conclusion: The proposed method performs better than automatic ellipse fitting. The higher accuracy and faster and more objective analysis suggest that the developed model is applicable in real-life scenarios, and can provide a valuable update to Bloodstain Pattern Analysis.
A Voxel-Wise, Cascaded Classification Approach to Ischemic Stroke Lesion Segmentation
Lecture Notes in Computer Science, 2016
Forensic Science International, 2015
1.2. A novel approach to automated bloodstain pattern analysis using an active bloodstain shape model
Journal of Forensic Radiology and Imaging, 2014
ABSTRACT Objectives Conventional blood pattern analysis is a tedious and time- consuming process ... more ABSTRACT Objectives Conventional blood pattern analysis is a tedious and time- consuming process due to the many actions that must be performed by the pattern analyst. Digital photographs can now be employed to aid in the process of scene analysis, by letting computer programs perform some of the required steps. However, current computerized methods are all based upon the assumption that any stain can be approximated by a simple ellipse, while manual work is still required. Materials and methods This work presents a novel approach, employing a regressed active shape model to approximate bloodstains. This Active Bloodstain Shape Model (ABSM) uses a regression to correlate the shape of a stain׳s approximation to its impact angle. This model is then deployed in a software pipeline, aiming to eliminate any user-input from the process. Fiducial markers are positioned in a crime scene, allowing for fully automated pattern analysis. Images are cleaned of any perspective distortions, after which stains are segmented and analyzed using said model. A robust principal component analysis (PCA) is then used to analyze the intersections of the reconstructed flight paths. Results Experimental results have demonstrated that the ABSM is able to better approximate stains, while accuracy on predicted impact angles is increased. Both simulated as well as real crimes scenes have confirmed the working of the automated pipeline. Conclusion Because the ABSM learns from experimental data, it is able to better approximate stains, resulting in an increased accuracy. The addition of the software pipeline removed the need for almost all user-input.
3.5. The feasibility of measuring body weight on CT images and the first steps in anatomical mirroring
Journal of Forensic Radiology and Imaging, 2014
ABSTRACT Objective To estimate body weight using the data of a post mortem total body CT scan and... more ABSTRACT Objective To estimate body weight using the data of a post mortem total body CT scan and establish its efficacy in a forensic context. Material and method PMCT data were used to compute fat, soft tissue and bone volume using in house developed software based on density and HU. The digitally calculated body weight was then compared with the measured body weight on autopsy. Results Analysis of the obtained results showed a good correlation between the measured body weight and the CT-derived body weight. The highest discrepancies were noticed in children and decomposed bodies presumably due to different bone mineralization in children versus adults and putrefactive gas formation and liquefaction in decomposed bodies. In general, there was no distinct consistent over- or underestimation. In case of partial scans, where a body part was not scanned or was missing, mirroring the unaffected side was seen to give equally good results. Conclusion The weight of a body can be sufficiently estimated using volume measurements of different body tissues obtained from the PMCT data. This technique can be applied when concern is raised about the documented body weight, or if weighing was omitted all together. Furthermore the aforementioned body weight estimates can be useful in case of mass casualties, for victim identification purposes, or in case of traumatic deaths (e.g. dismemberments).
Journal of Forensic Radiology and Imaging, 2014
Objectives: Traditional Bloodstain Pattern Analysis (BPA) estimates impact angles for bloodstain ... more Objectives: Traditional Bloodstain Pattern Analysis (BPA) estimates impact angles for bloodstain spatters by employing ellipses and an inverse sine. This approach is based upon a simplification of the stain formation process, and ignores the physical properties of blood, and its intricate interactions with the target surface. This research presents a data-driven approach, starting from experimental bloodstain spatters. No assumptions about the shape of stains or about the relation between the shape and their impact angle are made a priori. Materials and methods: Experimental bloodstain data is gathered, after which shape variations are extracted by employing an Active Shape Model. A non-linear regression is then used to explain the relation between stain shape and impact angle. Results: Experimental results show that traditional width-length ratios may deviate from the assumed inverse sine. The proposed approach, on the other hand, closely follows the experimental data through its regression. Additional experiments have shown an increased accuracy on impact angle estimates for the proposed approach when compared to automatic ellipse fitting. Conclusion: The proposed method performs better than automatic ellipse fitting. The higher accuracy and faster and more objective analysis suggest that the developed model is applicable in real-life scenarios, and can provide a valuable update to Bloodstain Pattern Analysis.
A Voxel-Wise, Cascaded Classification Approach to Ischemic Stroke Lesion Segmentation
Lecture Notes in Computer Science, 2016