Automated detection of regions of interest in cartridge case images using deep learning

This paper explores a deep-learning approach to evaluate the position of circular delimiters in cartridge case images. These delimiters define two regions of interest (ROI), corresponding to the breech face and the firing pin impressions, and are placed manually or by an image-processing algorithm. This positioning bears a significant impact on the performance of the image-matching algorithms for firearm identification, and an automated evaluation method would be beneficial to any computerized system. Our contribution consists in optimizing and training U-Net segmentation models from digital images of cartridge cases, intending to locate ROIs automatically. For the experiments, we used high-resolution 2D images from 1195 samples of cartridge cases fired by different 9MM firearms. Our results show that the segmentation models, trained on augmented data sets, exhibit a performance of 95.6% IoU (Intersection over Union) and 99.3% DC (Dice Coefficient) with a loss of 0.014 for the breech face images; and a performance of 95.9% IoU and 99.5% DC with a loss of 0.011 for the firing pin images. We observed that the natural shapes of predicted circles reduce the performance of segmentation models compared with perfect circles on ground truth masks suggesting that our method provide a more accurate segmentation of the real ROI shape. In practice, we believe that these results could be useful for firearms identification. In future work, the predictions may be used to evaluate the quality of delimiters on specimens in a database, or they could determine the region of interest on a cartridge case image.     

Development of etching paste for serial number restoration on aluminum engine block

Engine numbers, which involve information regarding the engine type, production number, and year and place of manufacture, are used for identification purposes. Comprising of unique alphanumeric characters, the engine numbers are fully or partially obliterated especially in auto theft and smuggling cases to conceal the origin, identity, and owner of vehicles. The limitations of the current restoration techniques such as the difficulty of using chemical liquid etching in vertical sites, the restrictions of magnetic and optical methods, and the applicability of several techniques like electron backscatter diffraction only in the laboratory environment prompt the development of new techniques. In view of these limitations and the importance of restoring engine numbers in criminal investigations, this unique study aimed to develop an etching paste that would restore the effaced characters on a real aluminum alloy engine block. The characters which were cold-stamped on the engine block were milled at varying depths and restoration attempts were conducted using etching pastes formed with different chemicals and materials. The analyses indicate that the etching paste formed with 200 mg of perlite, 400 mg of iron powder, and 450 μL of 20 M NaOH provided restoration to a good extent. The prevention of over-etching through the controlling of the chemical reaction and the cost-effectiveness appears to be the advantages of this technique. The success of recovery on the real engine block, the facilitation of restoration on curved surfaces, and the chance of on-site usage will likely make the etching paste a widely used tool in serial number restoration.     

Piercing the surface: A mechanical analysis of stabbing with household tools

Estimating the applied power during a stabbing incident, or estimating the minimal force necessary to penetrate the body with a certain weapon is a challenging task in forensic practice. A thorough forensic evaluation of stabbing forces needs objective numerical experimental data. Stabbing tests of 12 different weapons – including knives, a pair of scissors, a fork, screwdrivers, a rasp, a corkscrew, and a utility knife blade – were performed with a Mecmesin MultiTest-dV material tester on pork loin and ballistic gel to estimate the stabbing forces and dynamics. Penetration force (F_p) and maximal force (F_max) were recorded, and the registered force curves were analyzed. F_max was 159.8–212.07 Newton (N), 30.56–30.58 N, 168.9–185.48 N for various knives; 171.39–190.43 N for the pair of scissors, 233.6 N for the fork; 532.65–562.65 N, 370.31–367.19 N and 314.51–432.89 N for various screwdrivers, 44.14–56.62 N for utility knife during pork loin stabbing. The butter knife, corkscrew and rasp were not able to penetrate the pork loin, and the curved fork bent during stabbing. The results prove that weapon characteristics greatly influence the force necessary for penetration. Maximal stabbing force depends mostly on tip sharpness, and the force sharply decreases after penetration occurs, which indicates that edge sharpness is not as important as tip characteristics during stabbing perpendicular to skin surface. The penetration force during stabbing with a pair of scissors is comparable to the penetration force of knives. Stabbing with screwdrivers generally needs larger force than average knives but depends greatly on screwdriver size.     

Organic impurity profiling of fentanyl samples associated with recent clandestine laboratory methods

Nearly a decade ago, fentanyl reappeared in the United States illicit drug market. In the years since, overdose deaths have continued to rise as well as the amount of fentanyl seized by law enforcement agencies. Research surrounding fentanyl production has been beneficial to regulatory actions and understanding illicit fentanyl production. In 2017, the Drug Enforcement Administration (DEA) began collecting seized fentanyl samples from throughout the United States to track purity, adulteration trends, and synthetic impurity profiles for intelligence purposes. The appearance of a specific organic impurity, phenethyl-4-anilino-N-phenethylpiperidine (phenethyl-4-ANPP) indicates a shift in fentanyl production from the traditional Siegfried and Janssen routes to the Gupta-patent route. Through a collaboration between the DEA and the US Army's Combat Capabilities Development Command Chemical Biological Center (DEVCOM CBC), the synthesis of fentanyl was investigated via six synthetic routes, and the impurity profiles were compared to those of seized samples. The synthetic impurity phenethyl-4-ANPP was reliably observed in the Gupta-patent route published in 2013, and its structure was confirmed through isolation and structure elucidation. Organic impurity profiling results for illicit fentanyl samples seized in late 2021 have indicated yet another change in processing with the appearance of the impurity ethyl-4-anilino-N-phenethylpiperidine (ethyl-4-ANPP). Through altering reagents traditionally used in the Gupta-patent route, the formation of this impurity was determined to occur through a modification of the route as originally described in the Gupta patent.     

Fatal intoxication involving 2-methyl AP-237

Novel synthetic opioids contribute considerably to the opioid epidemic, especially with the frequent emergence of structurally similar compounds. This case report describes a fatal intoxication involving 2-methyl AP-237. A 54-year-old Caucasian male was found deceased from an apparent drug overdose. A plastic container labeled “2MAP” and a cut straw were found in the decedent's backpack at the scene. A white substance found in the container tested positive for fentanyl by field testing. According to his medical history, the decedent was treated for a drug overdose 3 years prior to his death. With no diagnostic findings at autopsy, the case was submitted for toxicological analysis. An unknown substance was detected in peripheral blood and urine using gas chromatography with nitrogen phosphorous detection (GC-NPD). Further testing was conducted using gas chromatography–mass spectrometry (GC–MS) and liquid chromatography-quadrupole-time-of-flight mass spectrometry (LC-QTOF-MS) which confirmed the presence of 2-methyl AP-237 and potential metabolites in blood and urine. Quantitation by GC-NPD revealed concentrations of 2-methyl AP-237 in blood and urine at 480 ng/mL and 4200 ng/mL, respectively. The toxicological analysis also identified and quantitated alprazolam in the blood at 55 ng/mL. Additionally, the metabolism of 2-methyl AP-237 was investigated and three hydroxylated metabolites were identified in peripheral blood and urine. Limited literature is available for the detection and quantitation of 2-methyl AP-237 in postmortem specimens. Given the toxicological findings with unremarkable autopsy findings, this case is an example of a fatal intoxication involving 2-methyl AP-237.     

2,4,6-Tri(2′-pyridyl)-1,3,5-triazine for determination of iron(II), Iron(III), and total iron contents on human palms

When iron-based tools, such as knives or guns, are held, traces of iron can transfer to the skin. However, no previous studies have been published regarding the effect of the elapsed contact time on the transfer of iron species with different valences to the palm. Compared with 3-(2-pyridyl)-5,6-diphenyl-1,2,4-triazine (PDT), 2,4,6-tri(2′-pyridyl)-1,3,5-triazine (TPTZ) was found to exhibit a higher sensitivity to iron(II) spectrophotometrically. This work employed 2,4,6-tri(2′-pyridyl)-1,3,5-triazine (TPTZ) and UV spectrophotometry to determine the amounts of iron(II), iron(III), and total iron transferred to human palms from iron tools. It was found that the palmar moisture level was an important factor in determining the amounts of total iron, including iron(II), transferred to the palm. For identical contact times, the amounts of total iron transferred to the palm was proportional to the palmar moisture, and the difference between the maximum and minimum amounts was 12 μg per hand. However, the amounts of iron(II) transferred to the palm gradually decreased over time for low palmar moisture levels, but steadily increased over time for high palmar moisture levels. Additionally, for average levels of palm moisture, the amounts of iron(II) and iron(III) transferred to the palm gradually decreased and increased, respectively, with longer contact times. Notably, this research could serve as a theoretical basis and guide for the detection of trace iron species with different valences on human palms for criminal investigations.     

Ballistic impact of hollow-point ammunition on porcine bone

Identifying failure mechanisms in skeletal tissue allows a deeper understanding of the effects of specific projectile impacts on bone. While ballistic trauma in flat bones is largely researched, knowledge of how long bones react to gunshot impacts is limited in the literature. The impacts of deforming ammunition appear to produce higher levels of fragmentation; however, these have not been studied in depth. This study compares the damage to femora bone by HP 0.357 and 9 mm projectiles constructed with both full and semi-metal jackets. Impact experiments were undertaken on a single-stage light gas gun involving the use of a high-speed video camera and full reconstruction of the bones to ascertain fracture patterns occurring in the femora. Higher degrees of fragmentation are likened to the presence of semi-jacketed HP projectiles than jacketed HP projectiles. The observations of external facing beveled edges are believed to be associated with the increased separation of the jacket and lead core of projectiles. Additionally, experimentation has shown that the amount of kinetic energy lost postimpact is likely related to the presence or the absence of a metal jacket on an HP projectile. The observed data, therefore, suggest that the composition, rather than the configuration, of a projectile affects the type and extent of the damage.     

A reference database of Windows artifacts for file-wiping tool execution analysis

Anti-forensic technology can play an effective role in protecting information, but it can make forensic investigations difficult. Specifically, file-wiping permanently erases evidence, making it challenging for investigators to determine whether a file ever existed and prolonging the investigation process. To address this issue, forensic researchers have studied anti-forensic techniques that detect file-wiping activities. Many previous studies have focused on the effects of file-wiping tools on $MFT, $LogFile, and $DATA, rather than on Windows artifacts. Additionally, previous studies that have examined Windows artifacts have considered different artifacts, making it difficult to study them in a comprehensive manner. To address this, we focused on analyzing traces in 13 Windows artifacts of 10 file-wiping tools' operations in the Windows operating system comprehensively. For our experiments, we installed each file-wiping tool on separate virtual machines and checked the traces that the tools left behind in each artifact. We then organized the results in a database format. Our analysis revealed that most of the tools left traces on other artifacts, except for JumpList, Open&SavePidlMRU, and lnk. There were also some cases where traces remained on the other three artifacts. Based on our research, forensic investigators can quickly identify whether a file-wiping tool has been used, and it can assist in decision-making for evidence collection and forensic triage.     

Pulp volume quantification methods in cone-beam computed tomography for age estimation: A critical review and meta-analysis

The secondary dentin deposition that occurs after the tooth formation process results in a decrease in pulp cavity volume in adolescents and adults. The purpose of this critical review was to correlate pulpal and/or dental volume on cone-beam computed tomography (CBCT) with chronological age approximation. A subobjective was to investigate which methodology and CBCT technical parameters would be most appropriate to evaluate this correlation. This critical review followed the PRISMA guidelines, and it was conducted by a search through PubMed, Embase, SciELO, Scopus, Web of Science, Cochrane Library databases, as well as gray literature. Primary studies that used pulp volume, or pulp chamber to tooth volume ratio measured using CBCT were included. Seven hundred and eight indexed and 31 non-indexed records were identified. A qualitative analysis was performed including 25 selected studies with a total of 5100 individuals, age ranging from 8 to 87 years without sex predilection. The most used method was pulp volume/tooth volume. CBCT voxel size ranged between 0.09 and 0.5. Manual segmentation associated with threshold algorithms was used in most of the studies. Correlation between the pulp volume/tooth volume ratio was moderate: −0.66 for the upper central incisors, −0.59 for upper canines and −0.56 for lower canines. High heterogeneity was observed among the studies. It is concluded that pulp volume should be used with caution in age estimation. Evidence supports the preferable use of upper incisors with pulp volume/tooth volume ratio for age estimation. There is not enough evidence that voxel size interferes in age estimation by pulp volume.     

Evaluation of canine training aids containment for homemade explosive and components by headspace analysis and canine testing

While canines are most commonly trained to detect traditional explosives, such as nitroaromatics and smokeless powders, homemade explosives (HMEs), such as fuel–oxidizer mixtures, are arguably a greater threat. As such, it is imperative that canines are sufficiently trained in the detection of such HMEs. The training aid delivery device (TADD) is a primary containment device that has been used to house HMEs and HME components for canine detection training purposes. This research assesses the odor release from HME components, ammonium nitrate (AN), urea nitrate (UN), and potassium chlorate (PC), housed in TADDs. Canine odor recognition tests (ORTs) were used with analytical data to determine the detectability of TADDs containing AN, UN, or PC. Headspace analysis by gas chromatography/mass spectrometry (GC/MS) with solid-phase microextraction (SPME) or online cryotrapping were used to measure ammonia or chlorine, as well as other unwanted odorants, emanating from bulk AN, UN, and PC in TADDs over 28 weeks. The analytical data showed variation in the amount of ammonia and chlorine over time, with ammonia from AN and UN decreasing slowly over time and the abundance of chlorine from PC TADDs dependent on the frequency of exposure to ambient air. Even with these variations in odor abundance, canines previously trained to detect bulk explosive HME components were able to detect all three targets in glass and plastic TADDs for at least 18 months after loading. Detection proficiency ranged from 64% to 100% and was not found to be dependent on either age of material.