Volume 24. Number 1
December 2025

Original Articles

Mass Spectrometric Differentiation and In Vitro Metabolic Profiling of Fluorodeschloroketamine Positional Isomers
Ya-Ling Yeh, M.Sc.; Yu-Wei Huang, M.Sc.; Che-Yen Wen, Ph.D.; Chin-Lin Hsieh, Ph.D.; Sheng-Meng Wang*, Ph.D.

DOI:10.6593/FSJ.202512_24(1).0001

Abstract：Ketamine and its analogs, such as 2-fluoro-deschloroketamine (2-FDCK), methoxetamine, deschloroketamine, and Bromodeschloroketamine, have garnered significant attention due to their widespread abuse and potential toxicity. Among these, 2-FDCK, 3-FDCK and 4-FDCK share structural similarities, yet only 2-FDCK is currently classified as a controlled substance in Taiwan. Given their differences in pharmacokinetics, pharmacodynamics, and metabolic pathways, the establishment of reliable analytical methods is crucial for forensic and toxicological investigations. In this study, we utilized Trifluoroacetic anhydride derivatization and the gas chromatography-mass spectrometry (GC-MS) to develop a mass spectral classification model for distinguishing these positional isomers. We used principal component analysis (PCA) for feature extraction and logistic regression (LR) for classification. In our experiments, the model achieved 100% accuracy in the test set. Analysis of urine samples confirmed the presence of 2-FDCK, demonstrating the method’s effectiveness in differentiating ortho-, meta-, and para-substituted isomers. Additionally, we conducted an in vitro metabolism study using human liver microsomes (HLMs) and liquid chromatography-quadrupole time-of-flight mass spectrometry (LC-QTOF-MS). This analysis identified 15 metabolites, formed through hydrogenation, N-demethylation, hydroxylation, deamination, and dehydration. Among the isomers, 4-FDCK metabolized the slowest, with 48% of the parent compound remaining and dihydro-4-FDCK as the major metabolite, while 3-FDCK exhibited the fastest metabolism, with dihydro-3-FDCK as the predominant product. Comparison with urine samples confirmed nor-2-FDCK (M1), dihydro-2-FDCK (M8(a)), and dihydro-nor-2-FDCK (M9(a), M9(b)) as key biomarkers for 2-FDCK detection, particularly when the parent compound is present at low concentrations. This study establishes a mass spectral classification model and an in vitro metabolite database, which facilitate rapid isomer differentiation and metabolite identification. The approach provides a valuable reference for forensic and toxicological applications, aiding in the detection of structurally similar new psychoactive substances.

Keywords: forensic Science, fluorodeschloroketamine (FDCK), isomer, in vitro metabolic study, machine learning analysis

The Research on the Trends, Challenges, and Misuse-Prevention of Deepfake Technologies
Wen-Chao Yang, Ph.D.*

DOI:10.6593/FSJ.202512_24(1).0002

Abstract：With the rapid advancement of artificial intelligence (AI), deepfake technologies—driven primarily by deep learning architectures, such as autoencoders, generative adversarial networks (GANs), and diffusion models—have enabled the creation of highly realistic synthetic audio-visual content. While these innovations have legitimate applications in entertainment and education, their misuse poses a serious risk to privacy, democracy, and national security. This study provides a comprehensive examination of the evolving threats, challenges, and preventive strategies associated with deepfake technology. It examines the impacts on media integrity, criminal justice, and public trust, alongside the emergence of legal and regulatory responses in Taiwan, the United States, the European Union, and China. Furthermore, the paper explores state-of-the-art deepfake detection and forensic methodologies, including metadata verification, frequency-domain analysis, noise profiling, and AI-assisted detection models. A standardized forensic workflow is proposed to enhance reproducibility and evidentiary reliability in judicial contexts. Ultimately, the study recommends a multifaceted framework that encompasses legal reform, technological innovation, public education, platform accountability, and international cooperation to mitigate the misuse of deepfake technologies while promoting the responsible development of AI.

Keywords: forensic science, deepfake, artificial intelligence (AI), legal regulation, digital evidence, information security

Evaluation of the Discriminatory Capacity in Commonly Used Y-STR Systems: Establishing a Reference for Forensic Interpretation
Chih-Wen Su*, Ph.D.; Chen-Ming Chen, M.S.

DOI:10.6593/FSJ.202512_24(1).0003

Abstract：In this study, two commonly used Y-chromosome short tandem repeat (Y‑STR) kits—the PowerPlex^® Y23 System and the Yfiler™ Plus PCR Amplification Kit—were validated and subjected to statistical analyses. We compared the discriminatory capacity (DC) of the kits: AmpFLSTR™ Yfiler™ (17 loci), DC = 0.897 (N = 970); PowerPlex^® Y23 (23 loci), DC = 0.978 (N = 418); Yfiler™ Plus (27 loci), DC ≈ 1 (N = 574); and for the 21 overlapping loci between PowerPlex^® Y23 and Yfiler™ Plus, DC = 0.955 (N = 970). These results indicate that the Y‑STR kits with more loci effectively enhanced the DC. We also observed that rapid‑mutation Y‑STR loci exhibited higher gene diversity (GD) and revealed no obvious differences between Taiwanese males and other groups. Because partial profiles are common in forensic DNA, we proposed a method for estimating the minimal number of Y‑STR loci sufficient for interpretation, which forensic laboratories could use under their guidelines. Ordering loci by ascending GD, we sequentially selected the lowest 8~23 loci (for PPY23) and 8~27 loci (for YFP27) and recomputed the number of unique haplotypes (UH) and DC. Using the first 12 loci, PPY23 yielded 321 UH among 418 samples (DC = 0.768; match probability [MP] = 0.0048; haplotype diversity [HD] = 0.998), whereas YFP27 yielded 423 UH among 574 samples (DC= 0.737; MP = 0.0047; HD = 0.997). These benchmarks provide practical guidance for reporting and interpreting evidentiary Y‑STR profiles.

Keywords: forensic science; Y-chromosomal short tandem repeat (Y-STR); discrimination capacity (DC); gene diversity (GD); match probability (MP); haplotype diversity (HD); sexual assault

Development of Latent Fingermarks Exposed to Simulated Destructive Environment using Powder Suspension
Hong-Jun Song, B.S.; Pu-Hsiang Chiu, B.S.; Wee-Beng Lim, M.S.; Wei-Ling Stella Tan, M.S.;Chun-Chieh Chen*, Ph.D.

DOI:10.6593/FSJ.202512_24(1).0004

Abstract：Recently, there has been increasing attention on improving the procedural fairness and scientific accuracy of crime-scene evidence collection to promote objective and impartial case analysis. Fingermarks, which can objectively link individuals to specific objects or places, remain essential evidence. On nonporous surfaces, the powder and cyanoacrylate fuming methods are commonly used to visualize fingermarks. However, they often perform poorly when exposed to harsh environmental conditions such as rain, deliberate submersion, or damage from fire, heat, and smoke. Thus, developing fingermark visualization methods on prints exposed to such challenging environments is essential. The wet powder suspension (WPS) technique—also called powder suspension—uses a liquid mixture containing solid powders and surfactants to visualize fingermarks. It includes the older small-particle reagent (SPR) and more recent, thicker paint-like formulations (Wetwop™ and Wet Powder™). The latter are effective on sticky surfaces such as tape. They can also be applied to other nonporous surfaces. In 2014, police in the United Kingdom recommended adding the WPS method to their Fingermark Visualization Manual. Owing to its simplicity and reliability, WPS performs effectively even under difficult conditions, including moisture or surface contamination. In this study, three commercial WPS reagents were tested on 11 nonporous substrates commonly found at crime scenes, including garbage bags, zip-lock bags, ceramic tiles, glass slides, rubber boots, polypropylene containers, aluminum foils, stainless-steel boxes, and gloves (PVC, nitrile, and latex). Four types of fingermark residues, namely natural, standard, sebaceous (oil-based), and amino acid-based prints, were deposited on these substrates before treatment with three reagents. Substrates that displayed good results will be exposed to simulated harsh environments before being processed for prints. ImageJ grayscale analysis was performed on the resulting images to evaluate for completeness and contrast in order to create objective evaluation criteria. Results from the experiment suggested that surface properties, residue composition, and environmental exposure do affect the clarity of fingermark ridges. The powder type also influences its adhesion and deposition on the surface for visualization. These findings under simulated destructive conditions provide practical recommendations for improving fingermark visualization success at real
crime scenes.

Keywords: forensic science, fingermark, wet powder suspension, destructive environment, water immersion, arson scene