Statistical Analysis of GSR Particles Morphometry Using the CART Method

Abstract

According to ASTM E1588-20, gunshot residue (GSR) particles can be unequivocally identified through chemical and morphometric analysis using scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDS), the gold standard technique for GSR detection. Recent studies have reported the presence of characteristic GSR particles—containing lead (Pb), barium (Ba), and antimony (Sb)—on vehicle occupants exposed to airbag deployment, underscoring the need for complementary analytical approaches. While elemental composition remains the primary criterion for GSR identification, morphometric analysis enhances the ability to differentiate GSR from other environmental particles. Furthermore, detailed characterization of GSR particle morphology may assist in determining the type of firearm used in a shooting incident. This study systematically analyzed characteristic GSR particles originating from four Brazilian-manufactured ammunition, establishing an initial framework for differentiating between two classes of firearms (short and long) based on morphometric features using the Classification and Regression Tree (CART) method. CART is well-suited for scenarios where interpretability and ease of implementation are priorities. Two short firearms—Taurus G2C pistol (0.40 caliber) and Glock G23 pistol (9 mm caliber) and two long firearms—Colt M16A2 rifle (5.56 mm caliber) and IMBEL FAL rifle (7.62 mm caliber) were tested: Ammunition types included CBC 0.40 S&W CSCV 160 gr, CBC 9 mm copper bullet (batch BNC10), CBC 5.56 mm AXO46 (batch A0142946), and CBC 7.62 × 51 mm Common. Morphometric analysis revealed distinct variations in characteristic GSR particle profiles across different ammunition calibers. A new four-category classification system for characteristic GSR particles was developed, with 57% identified as regular spheroids. Using CART analysis, a statistical model achieved 76% accuracy in distinguishing between short and long firearms based on morphometric parameters, particularly circularity and Feret diameter. Further research with expanded datasets and alternative predictive methods is recommended to enhance model performance and generalizability. These findings reinforce the potential of morphometric classification as a complementary tool in forensic ballistics.

Keywords

Automatization analysis / Classification / Firearm / Gunshot residue / Morphometry

Acknowledgments

The authors would like to express their sincere gratitude to FAPERJ (Research Support Foundation of the State of Rio de Janeiro) for the financial support that made this research possible. We also thank INMETRO (National Institute of Metrology, Quality and Technology) for providing the infrastructure and laboratory facilities essential to the development of this study. Special thanks are extended to our colleagues who contributed to the review and refinement of the English language in this manuscript.

Author Contributions

Conceptualization, R.C.S.; Methodology, A.L.M.d.S., C.B.d.P., M.A.R., I.d.A.R.; Software, A.L.M.d.S.; Data Curation, R.C.S., A.L.M.d.S., C.B.d.P.; Writing—Original Draft Preparation, A.L.M.d.S.; Writing—Review & Editing, R.C.S., C.B.d.P., I.d.A.R., M.A.M.G.; Visualization, R.C.S., I.d.A.R.; Supervision, R.C.S.

Ethics Statement

Ethical review and approval were waived for this study, as the research did not involve any experimental procedures with human participants. All data analyzed were fully anonymized prior to inclusion in the study, ensuring the protection of personal and sensitive information. Furthermore, the study was based exclusively on the routine collection of samples carried out by police authorities in the normal course of their activities, without any direct interaction with individuals or additional interventions beyond standard forensic practice.

Informed Consent Statement

Informed consent was waived for this study, as no experiments or interventions involving human participants were performed. All data were fully anonymized prior to analysis, ensuring the confidentiality and protection of personal information. Moreover, the study relied exclusively on forensic samples obtained through routine police procedures, without any additional collection or interaction beyond standard investigative practice.

Data Availability Statement

The data supporting this study’s findings are available from the corresponding author upon reasonable request.

Funding

We would like to thank funding support from Faperj—Research Support Foundation of the State of Rio de Janeiro, Brazil. Grant number: E-26/290.038/2021.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

[1] Basu S. Formation of Gunshot Residues. J. Forensic Sci. 1982, 27, 72-91. doi:10.1520/JFS11453J.

[2] SWGGSR,Guide for Primer Gunshot Residue Analysis by Scanning Electron Microscopy/Energy Dispersive X-ray Spectrometry 11-29-11. 2011. Available online: https://docs.wixstatic.com/ugd/4344b0_46da982f600840abbfa2100e44021875.pdf (accessed on 10 April 2024).

[3] ASTM International.Standard Practice for Gunshot Residue Analysis by Scanning Electron Microscopy/Energy-Dispersive Spectrometry. ASTM Standards, Designation E 1588-20. ASTM International: West Conshocken, PA, USA, 2020. Available online: https://store.astm.org/e1588-20.html (accessed on 10 April 2024).

[4] Laflèche DJ, Brière SJ, Faragher NF, Hearns NG. Gunshot residue and airbags: Part I. Assessing the risk of deployed automotive airbags to produce particles similar to gunshot residue. Canad. Soc. Forensic Sci. J. 2018, 51, 48-57. doi:10.1080/00085030.2018.1463202.

[5] Cox EP. A method of assigning numerical and percentage values to the degree of roundness of sand grains. J. Paleontol. 1927, 1, 179-183.

[6] Wadell H. Sphericity and roundness of rock particles. J. Geol. 1933, 41, 310-331.

[7] Barret PJ. The shape of rock particles, a critical review. Sedimentology 1980, 27, 291-303.

[8] Clark MW. Quantitative shape analysis: a review. J. Int. Assoc. Mathemat. Geol. 1981, 13, 303-320.

[9] Winkelmolen AM. Critical remarks on grain parameters, with special emphasis on shape. Sedimentology 1982, 29, 255-265. doi:10.1111/j.1365-3091.1982.tb01722.x.

[10] Barreiros FM, Ferreira PJ, Figueiredo MM. Calculating shape factors from particle sizing data. Part. Part. Syst. Character. 1996, 13, 368-373. doi:10.1002/ppsc.19960130607.

[11] Diepenbroek M, Bartholomä A, Ibbeken H. How round is round? A new approach to the topic ‘roundness’ by Fourier grain shape analysis. Sedimentology 1992, 39, 411-422. doi:10.1111/j.1365-3091.1992.tb02125.x.

[12] Igathinathane C, Pordesimo LO, Columbus EP, Batchelor WD, Methuku SR. Shape identification and particles size distribution from basic shape parameters using ImageJ. Comput. Electron. Agr. 2008, 63, 168-182. doi:10.1016/j.compag.2008.02.007.

[13] Blott SJ, Pye K. Particle shape: a review and new methods of characterization and classification. Sedimentology 2008, 55, 31-63. doi:10.1111/j.1365-3091.2007.00892.x.

[14] Takashimizu Y, Liyoshi M. New parameter of roundness R: circularity corrected by aspect ratio. Progr. Earth Planet. Sci. 2016, 3, 2. doi:10.1186/s40645-015-0078-x.

[15] Cruz-Matías I, Ayala D, Hiller D, Gutsch S, Zacharias M, Estradé S, et al. Sphericity and roundness computation for particles using the extreme vertices model. J. Comput. Sci. 2019, 30, 28-40. doi:10.1016/j.jocs.2018.11.005.

[16] Kara İ. The relationship between gunshot-residue particle size and Boltzmann distribution. Forensic Sci. Res. 2020, 7, 47-52. doi:10.1080/20961790.2020.1713433.

[17] Kara İ, Tahillioglu E. Digital image analysis of gunshot residue dimensional dispersion by computer vision method. Microsc. Res. Tech. 2022, 85, 971-979. doi:10.1002/jemt.23966.

[18] Madkour FA, Mohammed ES, Radey R, Abdelsabour‐Khalaf M. Morphometrical, histological, and scanning electron microscopic investigations on the hard palate of Rahmani sheep (Ovis aries). Microsc. Res. Tech. 2022, 85, 92-105. doi:10.1002/jemt.23887.

[19] Vermeij E, Duvalois W, Webb R. Koeberg M. Morphology and composition of pyrotechnic residues formed at different levels of confinement. Forensic Sci. Int. 2009, 186, 68-74. doi:10.1016/j.forsciint.2009.01.019.

[20] Tucker W, Lucas N, Seyfang KE, Kirkbride KP, Popelka-Filcoff RS. Gunshot residue and brakepads: Compositional and morphological considerations for forensic casework. Forensic Sci. Int. 2017, 270, 76-82.

[21] Kara L, Sarikavak Y, Lisesivdin SB, Kasap M. Evaluation of morphological and chemical differences of gunshot residues in different ammunitions using SEM/EDS technique. Environ. Forensics 2016, 17, 68-79.

[22] Wirth MA. Performance evaluation of image processing algorithms in CADe. Technol. Cancer Res. Treat. 2005, 4, 159-172. doi:10.1177/153303460500400207.

[23] De Martins de Souza AL, Carreteiro Damasceno J, Prado CB, Magalhães Gurgel MA, Carvalho Silva R. Uncertainty model for automated gunshot residue particle length measurements obtained from electron microscopy images. J. Microsc. 2023, 292, 105-116. doi:10.1111/jmi.13232.

[24] Dias FC. Use of ImageJ Software for Quantitative Analysis of Material Microstructure Images; INPE: São José dos Campos, Brazil, 2008; pp. 33-126.

[25] Hosokawa EO.Decision tree technique in data mining [Undergraduate thesis, Faculdade de Tecnologia de São Paulo]. Faculdade de Tecnologia de São Paulo. 2011. Available online: http://www.fatecsp.br/dti/tcc/tcc0003.pdf (accessed on 4 August 2023).

[26] Krishna S, Ahuja P. A chronological study of gunshot residue (GSR) detection techniques: a narrative review. Egypt. J. Forensic Sci. 2023, 13, 51. doi:10.1186/s41935-023-00369-8.

[27] Schwoeble AJ, Exline DL. Current Methods in Forensic Gunshot Residue Analysis, 1st ed.; CRC Press: Boca Raton, FL, USA, 2000. doi:10.1201/9781420042573.

[28] Garofano L, Capra M, Ferrari F, Bizzaro GP, Di Tullio D, Dell’Olio M, et al. Gunshot residue: further studies on particles of environmental and occupational origin. Forensic Sci. Int. 1999, 103, 1-21. doi:10.1016/S0379-0738(99)00035-3.

[29] Ditrich H. Distribution of gunshot residues—The influence of weapon type. Forensic Sci. Int. 2012, 220, 85-90. doi:10.1016/j.forsciint.2012.01.034.

[30] Chohra M, Beladel B, Baba Ahmed L, Mouzai M, Akretche D, Zeghdaoui A, et al. Study of gunshot residue by NAA and ESEM/EDX using several kinds of weapon and ammunition. J. Radiat. Res. Appl. Sci. 2015, 8, 404-410. doi:10.1016/j.jrras.2015.02.012.

[31] Brożek-Mucha Z. Chemical and morphological study of gunshot residue persisting on the shooter by means of scanning electron microscopy and energy dispersive X-ray spectrometry. Microsc. Microanal. 2011, 17, 972-982. doi:10.1017/S1431927611012141.

[32] Cook M. Gunshot residue contamination of the hands of police officers following start-of-shift handling of their firearm. Forensic Sci. Int. 2016, 269, 56-62.

[33] Papini CJ.Comparative Study of Methods for Determining Particle Size. Master Dissertation, Universidade de São Paulo, São Paulo, Brazil, 2003.

[34] Ferreira T, Rasband W. ImageJ User Guide IJ 1.46. 2012. 1. 89-91. Available online: https://imagej.net/ij/docs/guide/ (accessed on 30 July 2023).

[35] Wallace JS. Chemical Analysis of Firearms, Ammunition, and Gunshot Residue, 2nd ed.; CRC Press:Boca Raton, FL, USA, 2018; pp. 3-4.

[36] Mosteller F, Tukey JW. Data Analysis and Regression: A Second Course in Statistics, Reading; Addison-Wesley: Boston, MA, USA, 1977.

[37] Lima NM. Statistical Analysis of a Process Aimed at Increasing the Service Life of Starch Application Roller Coating in Paper Production. Undergraduate Thesis, Universidade Federal de São Carlos, São Carlos, Brazil, 2022.