Automatic speaker recognition (ASR) is a method used in forensic speaker comparison (FSC) casework. In FSC cases two voice recordings are compared to make an inference about speaker identity. When employing ASR for this task, the case audio is compared by software, resulting in a case score. This case score needs to be converted (calibrated) to a likelihood ratio, the formal metric of strength of evidence. An LR calculation set (calibration set), i.e. audio data representative of the case audio files, is needed to train a mapping between scores and likelihood ratios specific to the case. What audio can be deemed as representative of the case audio and what audio conditions to use as selection criteria is a subjective, case-specific decision by the forensic practitioner. This decision can be informed by research into the effect of audio conditions on the score-to-likelihood-ratio mappings and the resulting likelihood ratios. Using forensically relevant data, this type of research is done in this work. Multiple conditions are investigated in this manner, along with an analysis of the influence of number of speakers in the LR calculation set to serve as a baseline for variability. The results suggest duration differences below 40 s should be accounted for. Ethnolect differences, whether recordings are made in a car and speech durations above about 40 s appear not impactful. The same result is found for gender and the difference between operational data / lab data, except when cross-comparing those conditions.

Mass-fatality incidents involving severe thermal injury pose major challenges for disaster victim identification, particularly when access to the oral cavity is limited and conventional dental radiography cannot be performed. This study reports the first large-scale, real-world deployment of a portable cone-beam computed tomography (CBCT) system during the forensic identification of civilian victims following the October 7, 2023 attack in Israel and evaluates its contribution to odontological identification. We conducted an observational analysis of cases examined by the Israel Police forensic odontology teams, documenting body condition, availability of antemortem dental records, imaging modality used, and final identification outcomes. Portable CBCT was used primarily in victims with severe burns or restricted oral access. All scans were diagnostically interpretable, and repeat acquisition was required in only five cases. CBCT contributed directly to positive identification in 40 of 170 scanned victims (23%), all of which were later confirmed by DNA testing. Most identifications were based on dental treatment patterns, followed by dental morphology and jaw morphology. Brief hands-on training enabled rapid proficiency among forensic odontologists, allowing continuous use under field conditions. These findings demonstrate that portable CBCT is a practical, non-invasive imaging solution in mass-fatality settings, providing diagnostically useful data when conventional methods are not feasible and supporting timely forensic identification.

Bloodstains found at a crime scene must be accurately identified to correctly interpret the forensic context and events. Bloodstains (human or animal) sometimes appear on complex substrates such as sand or soil, which complicates their collection and analysis. Traditional presumptive tests for human blood may yield false positives in the presence of animal blood. This limitation, together with the need to preserve small stains, has motivated the use of non-destructive analytical techniques, such as vibrational spectroscopy. The aim of this work was to explore the combined use of ATR-FTIR, Raman microspectroscopy, and chemometrics for detecting and identifying human, dog, and cat bloodstains on sand and soil. ATR-FTIR and Raman spectral data were used to construct orthogonal partial least squares-discriminant analysis (OPLS-DA) models to assist in differentiation. In general, ATR-FTIR spectra combined with OPLS-DA demonstrated superior species discrimination capabilities compared with Raman spectra. Sex discrimination was further investigated using both datasets. However, the discriminatory performance of Raman spectroscopy was limited. These results highlight the potential of vibrational spectroscopy combined with chemometrics as a non-destructive approach for differentiating multispecies bloodstains on complex substrates. However, given the small number of donors included in this study, the findings should be considered preliminary, and further studies with larger sample sets are needed to confirm and extend these observations.

The rapid emergence of novel synthetic opioids, including fentanyl analogues and nitazene-class substances, represents a critical public health concern. Their high potency and continuous structural modification demand highly sensitive and adaptable analytical approaches for reliable forensic detection. This study evaluates dried blood spot (DBS) sampling as a robust alternative to conventional liquid matrices for the identification of synthetic opioids. A method based on liquid chromatography coupled with high-resolution mass spectrometry (LC-HRMS) was developed and validated. Data acquisition was performed in full-scan data-dependent acquisition (DDA) mode. The combined DBS-LC-HRMS methodology is suitable for the detection of known fentanyl analogues and nitazenes, while also allowing the identification of potentially unknown compounds through data analysis. The limits of detection (LOD) and quantification (LOQ) ranged from 0.05 to 0.2 ng/mL and 0.1-0.75 ng/mL, respectively. The method established a linear range of 0.5-3 ng/mL for all analytes, with validation results supporting the suitability of DBS as a sample storage and pre-treatment approach, with simplified collection, storage and transport compared to traditional matrices. Stability was also assessed, demonstrating that DBS is a more robust and cost-effective approach for analyte preservation than frozen whole blood over a 7 days period. Furthermore, the method was successfully applied to real samples, confirming its suitability as an effective tool for forensic toxicology screening and analysis. The proposed workflow demonstrated satisfactory analytical performance, enabling the identification and quantification of 37 synthetic opioids; Overall, this strategy provides a flexible and efficient tool for routine forensic toxicology, with the capacity to adapt to the dynamic and evolving landscape of illicit synthetic opioids.

This article introduces a semiotic-Bayesian epistemological framework for the analysis and interpretation of skeletal trauma in forensic anthropology. The model integrates Peircean semiotics with probabilistic inference to provide a structured, reproducible approach to evaluating forensic evidence under conditions of uncertainty. Bone injuries are conceptualized as signs that maintain a causal relationship with the events that produced them; their interpretation begins with abductive reasoning, through which the forensic anthropologist formulates plausible hypotheses about the origin of the observed trauma. These hypotheses are subsequently organized through the abductive-deductive-inductive (ADI) cycle and formally evaluated using likelihood ratios (LR), derived from empirical frequency data reported in the published literature, to quantify the strength of the evidence. This proposal is illustrated through three forensic cases involving cranial blunt force trauma, cervicocephalic injuries associated with interpersonal violence, and close-range gunshot wounds. In each case, the framework enabled logical quantification of evidential weight, enhancing interpretive transparency, mitigating cognitive bias and facilitating communication within judicial contexts. The methodology proposed bridges qualitative, context-driven interpretation with quantitative rigor, thereby addressing longstanding epistemological challenges aligning anthropological practice with modern legal standards of admissibility, and highlighting the important role of the LR as a logically coherent and reproducible measure of strength.

This study originates from a road traffic accident in which two vehicles collided in a curve. Alongside the physical evidence, location-related traces recovered from the Cache.sqlite database of an iPhone belonging to one of the drivers constituted the principal source of digital evidence. Beyond this specific case, analysis of these traces raised key forensic questions regarding their contribution to accident reconstruction. Based on these considerations, four investigative questions were defined, concerning the timing of the crash, phone manipulation before impact, speed estimation, and crash location. These questions were examined through a controlled experimental design involving 16 iPhone devices and high-precision Global Navigation Satellite System (GNSS) reference data, complemented by real casework observations. The results indicate that the moment of the crash can be approximated through speed analysis or spatial mapping of the recorded traces. Course values reflect the direction of motion rather than device orientation, thereby precluding their use for detecting phone handling. Speed values are reliable under steady-state conditions, but degrade substantially during braking. Regarding location, the analysis reveals three potential post-processing phenomena referred to as projection, correction, and anticipation. They can introduce significant positional errors, particularly around the moment of impact. These findings underline both the value and the limitations of iPhone location-related traces for forensic purposes. They further emphasise the need for a deeper understanding of their underlying creation mechanisms to ensure reliable interpretation in accident reconstruction.

Bitemark analysis has long been an important field within of forensic odontology, yet it remains controversial due to its inherent subjectivity and the limitations of currently adopted 2D methodologies. The National Institute of Standards and Technology has highlighted the lack of sufficient scientific foundation for conventional bitemark analysis, particularly with respect to bitemarks impressed on human skin, concerning uniqueness, reproducibility, and evidentiary reliability. Rather than addressing the limitations identified for human skin bitemark analysis, this study proposes a three-dimensional quantitative protocol specifically designed for the analysis of bitemarks impressed on a standardized food substrate. Bitemarks were imprinted on standardized Italian cheese (scamorza) by eight adult volunteers. Dental casts and corresponding food items were digitized using a REVOPOINT Blue Light Mini 3D scanner. Digital models were processed with GOM Inspect and analyzed through: surface superimposition with colorimetric deviation mapping, extraction of 2D profiles from 3D data, and computation of three quantitative indicators: three-point angle, distance-perpendicular ratio, and elliptical ratio. A probabilistic model evaluated the likelihood of a match between each bitemark and dental arch. The 3D workflow minimized evidence handling and preserved bite pattern integrity. Colorimetric maps highlighted deviations within 5 mm and 2 mm tolerance ranges, enabling screening and refined matching. Quantitative indicators showed good discriminatory power, with correct matches for all subjects except one. Substrate deformation cases emphasized integrating surface analysis with profile-based indicators to reduce geometric distortion. The combined use of 3D scanning and reverse engineering reduced angular distortion and information loss typical of 2D methods, yielding more consistent and reproducible results within the experimental conditions adopted in this study. This study demonstrates a standardized engineering-based workflow for the quantitative analysis of bitemarks impressed on a stable food substrate. The proposed protocol provides objective and reproducible quantitative parameters for this specific experimental scenario and should not be interpreted as validation or direct support for bitemark analysis performed on human skin, whose biomechanical characteristics and distortion mechanisms differ substantially from those of stable inanimate substrates.

Ketamine has legitimate medical use in the clinic as an analgesic, with the additional recent approval of the single ketamine isomer, (S)-ketamine/esketamine, as a therapy for treatment resistant depression. Unfortunately, ketamine is also prone to non-medical use, often as a club or party drug. Herein, we describe the development and validation of a ketamine enzyme immunoassay screen, and both achiral and chiral LC-MS/MS confirmation methods. Our screen cutoff was set at 500 pg ketamine/mg hair. The confirmation cutoff, following an extended 90% ethanol wash and drug extraction, was set at 500 pg ketamine/mg hair, with the additional requirement of 50 pg norketamine metabolite/mg hair. Achiral method linearity was obtained from 10 to 5000 pg ketamine and norketamine/mg hair, chiral method linearity was obtained from 50 to 5000 pg (R)-ketamine and (S)-ketamine/mg hair. The LOD and LLOQ were set at 10 pg analyte/mg hair for the achiral method and 50 pg analyte/mg hair for the chiral method. The LC-MS/MS confirmation method was applied to the analysis of n = 18 authentic hair samples containing ketamine from the United States workplace drug testing population.

Stature estimation is one key component of the biological profile in forensic anthropology, yet most available regression formulae are derived from non-contemporary and non-UK reference samples. This study aimed to develop and evaluate preliminary PMCT-derived regression equations for estimating cadaveric length from long-bone measurements in a contemporary East London mortuary sample; and to compare their performance with commonly used Trotter and Gleser formulae. Whole-body PMCT scans of 100 adults from the East London Forensic Centre were analysed, comprising 50 males and 50 females aged 25-97 years. Bilateral measurements of the humerus, femur, tibia, and fibula were recorded in Horos using standard osteometric definitions. Side-averaged values were used in ordinary least squares regression models, with cadaveric length as the outcome variable. Both univariate and limited multivariate models were generated for pooled and sex-specific samples, and model performance was assessed in a number of ways such as adjusted R[2], standard error of estimate, root mean square error, mean absolute error, and cross-validation. Results indicate that lower limb variables showed the strongest associations with cadaveric length. However, the high inter-predictor correlations and VIF values indicated severe multicollinearity, meaning that the most complex multivariable models should not be interpreted biologically or treated as the strongest practical option. Instead, the findings support the use of parsimonious, sex-specific equations, with wider multivariable models presented cautiously as internally fitted models only. This study provides population-relevant PMCT-derived reference equations for a small contemporary East London sample, but external validation is required before routine forensic implementation.

Strangulation is a form of violence that involves the application of blunt force to the neck to control, coerce, or incapacitate. Most strangulation detection occurs within the soft tissues of the head and neck; however, robust osteological understanding is essential to ascertain the involvement of strangulation in those who are alternatively preserved, decomposed, or skeletonized, where soft tissue findings are unavailable or unreliable. The study analyzed head and neck injuries from 120 cases of fatal interpersonal strangulation recorded within the National Coronial Information System (NCIS) Database in Australia. The study aimed to augment our current understanding of head and neck injury patterns resulting from fatal interpersonal strangulation and provide a formalized Australian perspective on the phenomenon of interpersonal strangulation. Injuries were documented in almost all areas of the head and neck, with the most common injuries occurring in soft tissue structures such as muscle and skin. Some variables, such as sex, alcohol use, age, and BMI, were found to be statistically significantly associated with the odds of injury in specific anatomical regions, including cutaneous injuries of the neck, neck musculature injuries, intraoral space injuries, and laryngeal cartilaginous skeleton fractures. The findings contribute to our current understanding of strangulation in Australia, encompassing pathological, odontological, and anthropological perspectives, and provide exploratory statistical analysis and descriptive epidemiology of injury patterns relevant to medico-legal investigation.