Comparison of feature selection and data fusion of Fourier transform infrared and Raman spectroscopy for identifying watercolor ink

The identification of different kinds of watercolor inks is an important work in the field of forensic science. Four different kinds of watercolor ink Spectroscopy data fusion strategies (Fourier Transform Infrared spectroscopy and Raman spectroscopy) combined with a non-linear classification model (Extreme Learning Machine) were used to identify the brand of watercolor inks. The study chose Competitive Adaptive Reweighted Sampling (CARS), Random Frog (RF), Variable Combination Population Analysis-Genetic Algorithm (VCPA-GA), and Variable Combination Population Analysis-Iteratively Retains Informative Variables (VCPA-IRIV) to extract characteristic variables for mid-level data fusion. The Cuckoo Search (CS) algorithm is used to optimize the extreme learning machine classification model. The results showed that the classification capacity of the mid-level fusion spectra model was more satisfactory than that of single Infrared spectroscopy or Raman spectroscopy. The CS-ELM models based on infrared spectroscopy used to recognize the watercolor ink according to brands (ZHENCAI, DELI, CHENGUANG, and STAEDTLER) obtained an accuracy of 66.67% in the test set using all spectral datasets. The accuracy of CS-ELM models based on Raman spectroscopy was 67.39%. The characteristic wavelength selection algorithms effectively improved the accuracy of the CS-ELM models. The classification accuracy of the mid-level spectroscopy fusion model combined with the VCPA-IRIV algorithm was 100%. The data fusion method increased effectively spectral information. The method could satisfactorily identify different brands of watercolor inks and support the preservation of artifacts, paintings, and forensic document examination.     

A study of DNA transfers onto plastic packets placed in personal bags

The ability to detect low level DNA brings with it the uncertainty of whether the detected DNA is a result of transfer. To address this uncertainty, a simulation study was conducted in which a mock illicit drug packet was placed into the personal bags of individuals. When the average transit time of the packets was increased from around 2 h to more than 14 h, the percentage of the DNA profiles recovered from the packets which could be attributed to the individuals increased greatly from 5.3% to 48.6%. We found that drug packers who were poor shedders could not be included as contributors to the DNA profiles from the drug packets at all and there was a higher chance that individuals other than themselves could be included as contributors to the DNA profile recovered from drug packets. We also found that it was equally likely that the drug packers who had direct contact with the drug packets and bag owners who did not, could be included as contributors to the DNA profiles recovered from the packets. The results in this study highlight the importance of taking into consideration the transit time of drug packet, the shedder status of the alleged packer and the history of an item, when evaluating DNA evidence in the context of illicit drug activities.     

A novel chitosan/tripolyphosphate/L‐lysine conjugates for latent fingerprints detection and enhancement

Most chemical and physical methods employed in visualizing latent marks have shown detrimental effects on human health and, therefore, some research groups have directed their attention to the utilization of various (bio)polymers with the aforementioned purpose. Although chitosan is widely used in medicine, pharmacy, food, and drug delivery systems, there are very few studies that address this biopolymer utilization in forensic applications, such as the detection of latent fingerprints. We used chitosan crosslinked with sodium tripolyphosphate and conjugated with L‐lysine to enhance the visualization of latent fingerprints, due to its ability to interact with fingerprint sweat residues. These conjugates were prepared using different (w/w) amounts of chitosan and tripolyphosphate (6/1; 4/1; 1/1; 1/4; and 1/6), and those with the most promising results (i.e., 6/1 formulation) were investigated in detail. Fourier transform infrared (FT‐IR) spectroscopy confirmed interactions between components of the systems. Optical microscopy and scanning electron microscope (SEM) analysis showed that prepared powder formulations were uniform in size and confirmed that chitosan/tripolyphosphate/lysine conjugates bind easily to the sweat and lipid residues present in the latent fingerprints. The testing of prepared conjugates demonstrated the potential of these systems as bio‐based powder substitution for commercially available powders.     

Rapid and nondestructive analysis of lipstick on different substrates using ATR-FTIR spectroscopy and chemometrics

Lipstick can be an important piece of evidence in crimes like murders, rapes, and suicides. Due to its prevalence, it can be an important corroborative evidence in crime reconstruction. The analysis of such evidence can provide an evidentiary link between the suspect, the victim, object, or the crime scene. We report the use of nondestructive ATR-FTIR spectroscopy combined with chemometrics for the classification of 10 brands of lipsticks with nine samples each. Chemometric method of partial least square-discriminant analysis (PLS-DA) has been employed to interpret the data and classify the samples into their respective classes. The PLS-DA model provides an AUC figure above 0.99 in all brands except one; for which it is slightly less at 0.94. We have also tested the traces of these lipstick samples on different substrates treating them as unknowns in the already trained PLS-DA model. 100% of the samples on nine substrates viz. a cotton, nylon, plastic, dry tissue, denim (blue jeans), wet tissue, nitrile gloves, white paper, and polyester were correctly attributed to their source brand. In conclusion, the results suggest that ATR-FTIR combined with the chemometrics is a rapid, nondestructive, and accurate method for the discrimination and source attribution of lipstick. This study has potential for use in actual forensic casework conditions.     

Bruise dating using deep learning

The bruise dating can have important medicolegal implications in family violence and violence against women cases. However, studies show that the medical specialist has 50% accuracy in classifying a bruise by age, mainly due to the variability of the images and the color of the bruise. This research proposes a model, based on deep convolutional neural networks, for bruise dating using only images, by age ranges, ranging from 0–2 days to 17–30 days, and images of healthy skin. A 2140 experimental bruise photograph dataset was constructed, for which a data capture protocol and a preprocessing procedure are proposed. Similarly, 20 classification models were trained with the Inception V3, Resnet50, MobileNet, and MnasNet architectures, where combinations of learning transfer, cross‐validation, and data augmentation were used. Numerical experiments show that classification models based on MnasNet have better results, reaching 97.00% precision and sensitivity, and 99.50% specificity, exceeding 40% precision reported in the literature. Also, it was observed that the precision of the model decreases with the age of the bruise.     

Forensic Use of A Subtropical Blowfly: The First Case Indicating Minimum Postmortem Interval (mPMI) in Southern Brazil and First Record of Sarconesia Chlorogaster from a Human Corpse

Southern Brazil is unique due to its subtropical climate. Here, we report on the first forensic entomology case and the first record of Sarconesia chlorogaster (Wiedemann) in a human corpse in this region. Flies' samples were collected from a body indoors at 20°C. Four species were found, but only Chrysomya albiceps (Wiedemann) and S. chlorogaster were used to estimate the minimum postmortem interval (mPMI). The mPMI was calculated using accumulated degree hour (ADH) and developmental time. The S. chlorogaster puparium collected was light in color, so we used an experiment to establish a more accurate estimate for time since initiation of pupation where we found full tanning after 3 h. Development of C. albiceps at 20°C to the end of the third instar is 7.4 days. The mPMI based on S. chlorogaster (developmental time until the third instar with no more than 3 h of pupae development) was 7.6 days.     

Geoforensic search to crime scene: Remote sensing,geophysics, and dogs

In the absence of surface indications of burial sites, law enforcement or humanitarian organizations are faced with the difficult task of focusing large-scale ground searches to a manageable excavation area. A geoforensic-based survey may exclude parts of the landscape for reasons such as diggability or viewshed analysis but leave areas still too large for invasive exploration. This work examines how drone-based remote sensing, geophysics, and search dogs may be combined to narrow such searches. Here, we ask the reader to consider two examples where forensic geomorphology and land use provided a range of possible burial locations. Following this is a multi-proxy approach to similar dilemma, with a search-to-scene case study using remote sensing (drone photography), geophysics, ground probes, and search dogs. This approach is not presented as a definitive guide, but serves as an example of the conjunctive use of well-studied methods to approach a common problem in geoforensics.     

Stability studies of ALD-52 and its homologue 1P-LSD

With the emergence of new psychoactive substances (NPSs) over the years, the substances detected on stamps (also known as blotter papers) have also evolved from the traditional drug—lysergic acid diethylamide (LSD) to the multiple variants of lysergamides such as ALD-52 and 1P-LSD. The analysis of such blotter papers is usually done by solvent extraction followed by identification using gas chromatography–mass spectrometry (GC-MS). This study has shown that hydrolysis to form LSD was observed in GC-MS analysis when ALD-52 was extracted with methanol. The extraction of ALD-52 using other solvents such as acetonitrile, ethanol, isopropyl alcohol, ethyl acetate, and acetone, followed by GC-MS analysis, was investigated. It is shown that alcoholic solvents such as methanol and ethanol will result in the conversion of ALD-52 to LSD during GC-MS analysis, whereas the sterically hindered isopropyl alcohol will prevent this conversion. Investigation also shows that the hydrolysis of ALD-52 to LSD occurs at the GC injector port. It was also observed that the degree of hydrolysis was more pronounced at a lower concentration (0.1 mg/mL). The study was extended to a close analog—1P-LSD, and the results showed that 1P-LSD similarly hydrolyzes to LSD. However, 1P-LSD was observed to be more stable than ALD-52 due to steric hindrance because of the propanoyl group.     

Deep learning for sex determination: Analyzing over 200,000 panoramic radiographs

The objective of this study is to assess the performance of an innovative AI-powered tool for sex determination using panoramic radiographs (PR) and to explore factors affecting the performance of the convolutional neural network (CNN). The study involved 207,946 panoramic dental X-rays and their corresponding reports from 15 clinical centers in São Paulo, Brazil. The PRs were acquired with four different devices, and 58% of the patients were female. Data preprocessing included anonymizing the exams, extracting pertinent information from the reports, such as sex, age, type of dentition, and number of missing teeth, and organizing the data into a PostgreSQL database. Two neural network architectures, a standard CNN and a ResNet, were utilized for sex classification, with both undergoing hyperparameter tuning and cross-validation to ensure optimal performance. The CNN model achieved 95.02% accuracy in sex estimation, with image resolution being a significant influencing factor. The ResNet model attained over 86% accuracy in subjects older than 6 years and over 96% in those over 16 years. The algorithm performed better on female images, and the area under the curve (AUC) exceeded 96% for most age groups, except the youngest. Accuracy values were also assessed for different dentition types (deciduous, mixed, and permanent) and missing teeth. This study demonstrates the effectiveness of an AI-driven tool for sex determination using PR and emphasizes the role of image resolution, age, and sex in determining the algorithm's performance.     

Quantitative PCR overestimation of DNA in samples contaminated with tin

Metals can pose challenges while conducting forensic DNA analysis. The presence of metal ions in evidence-related DNA extracts can degrade DNA or inhibit PCR as applied to DNA quantification (real-time PCR or qPCR) and/or STR amplification, leading to low success in STR profiling. Different metal ions were spiked into 0.2 and 0.5 ng of human genomic DNA in an “inhibition study” and the impact was evaluated by qPCR using the Quantifiler™ Trio DNA Quantification Kit (Thermo Fisher Scientific) and an in-house SYBR Green assay. This study reports on a contradictory finding specific to tin (Sn) ions, which caused at least a 38,000-fold overestimation of DNA concentration when utilizing Quantifiler Trio. This was explained by the raw and multicomponent spectral plots, which indicated that Sn suppresses the Quantifiler Trio passive reference dye (Mustang Purple™, MP) at ion concentrations above 0.1 mM. This effect was not observed when DNA was quantified using SYBR Green with ROX™ as the passive reference, nor when DNA was extracted and purified prior to Quantifiler Trio. The results show that metal contaminants can interfere with qPCR-based DNA quantification in unexpected ways and may be assay dependent. The results also highlight the importance of qPCR as a quality check to determine steps for sample cleanup prior to STR amplification that may be similarly impacted by metal ions. Forensic workflows should recognize the risk of inaccurate DNA quantification of samples that are collected from substrates containing tin.