本文探讨 DNATyper mtDNA-SNP60 TM 试剂盒在案件中应用的可行性。应用 DNATyper mtDNA-SNP60^TM试剂盒对 100 个汉族无关个体和 20 组全同胞进行 mtDNA SNP 检验；取 25 pg/μL 马、牛、羊、猪、鸡、鸭、猫、狗、兔、鼠和大肠杆菌的 DNA 样品进行种属特异性测试；取 5、10、20、40 μmol/L 血红素进行抗抑制性测试；取两个批次的 DNATyper mtDNA-SNP60 TM 试剂盒经反复冻融 10 次后进行稳定性测试；分别应用 VeriFiler TM Plus PCR 扩增试剂盒和 DNATyper mtDNA-SNP60 TM 试剂盒对 100 份陈旧、腐败、降解检材进行检验。结果表明，100个汉族无关个体均获得清晰的 mtDNA SNP 分型结果，其检验结果与通过 mtDNA 测序获得的结果完全一致；100 个汉族无关个体含有 100 种不同的单倍型，20 组全同胞中每组个体之间 mtDNA SNP 分型结果相同；DNATyper mtDNA-SNP60 TM 试剂盒对马、牛、羊、猪、鸡、鸭、猫、狗、兔、鼠和大肠杆菌的 DNA 样品进行检测，均未出现特异性分型，表明该试剂盒具有较好的种属特异性；当血红素浓度≤40 μmol/L 时，所有 mtDNA SNP 位点均获得正确分型，表明该试剂盒具有一定的抗血红素能力；两个批次的 DNATyper mtDNA-SNP60 TM 试剂盒经反
复冻融 10 次后，所有 mtDNA SNP 位点均可正确分型，表明该试剂盒具有较好的稳定性；对于 100 份陈旧、腐败、降解检材，STR 检出率为 55%，mtDNA SNP 的检出率为 86%，mtDNA SNP 的检出率显著高于 STR。当模板 DNA 浓度大于 5 pg/μL 时，DNATyper mtDNA-SNP60 TM 试剂盒能得到完整的分型谱图。综上，DNATyper mtDNA-SNP60 TM 试剂盒可应用于陈旧、腐败、降解检材的检验，具有很好的实战应用价值。

In clinical medicine, the material density analysis function of energy spectrum technology takes water-bone-based material equivalence as the reference material for comparative analysis, which is mainly used for material quantitative and qualitative analysis. In the daily work of forensic medicine, the changes of lung tissue of drowning corpses are of great research value. Therefore, the virtual anatomy technique based on energy spectrum is used to analyze the characteristic changes of the lungs of rabbit drowning model, and to explore its role in the identification of drowning. Three death models of drowning, postmortem water intake and control group were respectively established for rabbit. Energy spectrum CT scan was performed on the experimental animals, and the differences in imaging features between the groups were compared by morphological reading of the tomographic images. Energy spectrum CT data were imported into the CAMPO IMAGING system. Peak analysis-energy spectrum display function to get the base material pair image and value, calculate the lung CT value; analyze the difference of lung base material value and CT value of each experimental group. The CT images of lungs in the drowning group showed characteristic ground glass changes. No obvious abnormality was found in the corresponding parts of the postmortem water intake group and the control group. The water volume of water-bone base material and CT value in drowning group were significantly higher than postmortem water intake group and control group (P<0.05). On the basis of the lung images of drowning, the image and numerical changes of water-bone base materials of energy spectrum CT combined with CT values can effectively reflect the characteristic changes of virtual anatomy of drowning lungs and provide a new reference for the identification of drowning. In the next experiment, the energy spectrum CT will be used to analyze the fluid state in the cardiac cavity of the drowning, so as to study the imaging characteristic changes. At the same time, it will also summarize the research on the mutual use of energy spectrum CT and conventional CT, in order to draw a more accurate research conclusion. However, there were some limitations in this study, only typical drowning been studied, and only animal experiments been used, and there was still a lack of human experiment; in addition, the corrupt gas produced in the corpse with the extension of the time of death may affect the scanning imaging. It is still a long way from the actual promotion, but the experimental results show the feasibility of the research direction, which will be supplemented and improved in the follow-up research.

Heat-induced fluorescence technology is a professional technology that can develop latent fingerprints on porous substrates without chemical reagents. Developing latent fingerprints on painted walls is a necessary application extension of this technology. Therefore, this paper focuses on exploring the reaction mechanism of the heat-induced fluorescence method of the latent fingerprint on the painted wall. On this basis, taking the painted wall as the object of trace-bearing, the influence of the heating distance, time length, temperature, coating type, and the aging degree of the latent fingerprint on the developing effect was gradually studied by using the control variable method. Finally, the best conditions of the latent fingerprint on the painted wall were obtained as follows: the heating distance of 15 cm, the heating time of 5 minutes, and the heating temperature of 500℃; the type of coating does not affect the effect of the heat-induced fluorescence technology to develop the latent fingerprints on the wall, and the long-dated latent fingerprints still have a good result. Technical research can provide technical support for developing latent fingerprints on the wall at the scene of criminal cases.

For the analysis of the suspected drugs seized in a local case, and for the identification and characterization of the compound, this paper identified and characterized the target compound based on several techniques including gas chromatography-mass spectrometry (GC-MS), liquid chromatography-high resolution time-of-flight mass spectrometry (LC-QTOF), Fourier transform infrared spectroscopy (FT-IR), hydrogen nuclear magnetic resonance spectroscopy (¹H NMR), carbon nuclear magnetic resonance spectroscopy (¹³C NMR) and X-ray diffraction (XRD) of single crystal. The fragment information and neutral loss mass spectrum of ketamine and F-ketamine were determined and compared with the target compound. Based on the characterization results and neutral loss mass spectrum, the compound possesses bromine stable isotopes and a similar structure with ketamine and F-ketamine, which has been identified as Br-ketamine. Ketamine, one of the most widely abused drugs, and its derivative F-ketamine (2-FDCK) have been both forbidden by the Chinese government. To avoid the legal sanction, drug criminals usually modify the chemical structures of old drugs. It can be expected that Br-ketamine has some risks of drug abuse. However, there was no relevant research on characterization, psychoactivity, addiction and physiological toxicity. Therefore, this work provides identification and characterization of Br-ketamine, a new phencyclidine derivative. On one hand, this work provides significant references for the identification of Br-ketamine and further research. On another hand, the proposed identification method using neutral loss mass spectra can also be applied to other unknown compounds.

Single nucleotide polymorphism (SNP) profiling is a commonly used genetic tool for individual identification and ancestry inference in forensic genetics. This study was constructed based on the whole genome SNP data of 163 Mongolian individuals from Inner Mongolia region of China and 777 individuals from 13 other populations in China. PLINK and ADMIXTURE software was used to analyze the genetic structure of these populations. Plink software was used for Genome Wide Association Studies to screen Mongolian specific SNPs. The DAA software was used to calculate the population matching probability, likelihood ratio and ancestral components of the reference population and 569 individuals from 7 test populations to evaluate the accuracy of the inference model. The results of population genetic structure analysis showed that the Mongolian population is mainly composed of northern East Asian component, and is closely clustered with other Altaic populations. 644 Mongolian specific AISNP loci were screened. Among the reference populations, the inference accuracy of the Mongolian population was 98.16%, and the other populations were 100%. Among the test populations, the accuracy of the Mongolian population was 86.67%, and the other populations were 96.15% to 100%. This study indicates that there are genetic structure differences between the Mongolian population and other East Asian subpopulations. The 644 selected AISNP loci can be used for genetic inference between the Mongolian population and other East Asian subpopulations, which is of great significance for improved ancestry inference within East Asia.

Paralytic shellfish poisoning is globally among the most commonly-seen poisoning incidents resulted from shellfish, posing huge threat to people’s life and health with its widest distribution, greatest harm and highest mortality (its oral-lethal dose less than 1mg for human). Paralytic shellfish poisoning is caused from the highly-toxic paralytic toxins of shellfish. The paralytic shellfish toxins (PSTs) are able to paralyze nerves and muscles and affect both the nerval and cardiovascular systems. PSTs have many kinds and structure-similar derivatives that can transform into one another, hence conducing difficulties to detect them. A review was here carried out into the PSTs about their properties, mechanism of action, toxic symptoms and metabolic pathways, having summarized the relating pretreatment methods of solid-phase extraction, dispersive solid-phase extraction and immunoaffinity chromatography, and successively discussed the applicable detection choices including the biologically-infected mouse approach, immunoassay, liquid-chromatography fluorescent detection and liquid chromatography-mass spectrometry. Therefore, valuable references should be capable of providing with this review for the concerned people to research PSTs, develop and improve relevant detection performance, examine correlative forensic-significant toxins and stipulate involving norms and standards.

The paper aims to study the occurrence of close non-matches (CNMs) in large-scale databases to enhance the risk awareness of fingerprint practitioners when they compare the suspect’s fingerprint against the ones from AFIS. Based on the database of more than 10 million people, the central areas of 40 arches were selected for research. Each fingerprint is marked with 40 minutiae to start sending queries. Reducing the number of feature points, in turn, to explore whether the amount can influence the AFIS occurrence rate of homologous fingerprints and CNMs, and make statistics on the similar areas of fingerprint and patterns of CNMs. Usually, as the number of minutiae increases, the occurrence rate of homologous features increases, and the occurrence rate of CNMs decreases. A total of 454 CNMs were found in 40 arches of volunteers’ fingerprints, one of CNMs was completely consistent with 15 feature paints. Average per query, the occurrence rate of CNMs in the top 100 candidates of each query is about 75.7%, of which the proportions of CNMs of the low, medium, and high levels are 46.5%, 19.7%, and 9.5%, respectively. In addition, CNMs appeared more in the arch pattern center’s 3, 4, and 6 regions, with an occurrence rate of 44.27%, 24.45%, and 18.5%, respectively, and CNMs were more from the triangular region of the dustpan pattern. This study contributes to improving the ability of fingerprint practitioners to identify CNMs, suggesting that fingerprint practitioners should pay more attention to the occurrence of CNMs in the middle-under parts of the central area of the arch.

In the investigation of telecom fraud dens, it is a key link to investigate the electronic equipment involved and extract electronic data, and the on-site evidence collection of memory data is an essential part of it. The scene of telecom fraud dens is usually complex and diverse, which commonly does not have the conditions to use large-scale forensics equipment. At the same time, there are numerous types and quantities of equipment involved, and the time for forensics is tight, which brings great technical difficulties and challenges to the scene forensics of memory data. Based on the relevant principles in “General Method for Collecting Electronic Evidence Data on the Spot”, this paper studied the current situation of memory data extraction in on-site evidence collection of telecom fraud dens. Firstly, according to the actual needs of the scene investigation of telecom fraud dens, the memory extraction methods of Windows, MAC and Linux systems in normal state were introduced. Then, for some troublesome problems, from the mechanism of operating system hibernation and unexpected system crash, the influence of hibernation files and dump files on memory extraction was analyzed, and the memory extraction method without knowing the password was proposed by using DMA dynamic memory reading technology. Finally, based on the characteristics of memory extraction in den investigation, a scheme of scene memory forensics was proposed and applied to practical cases, which provided reference for the exploration and evidence collection of such cases.

Detecting and identifying the combustion promoter residue in the fire site sample is a difficult task. The fire combustion process may be accompanied by pyrolysis, weathering/evaporation or microbial degradation of the combustion agent, etc., making the collection of residues difficult, and the subsequent identification of the fire combustion agent material evidence also brings interference, so it is very important to determine the nature of the fire and find out the cause of the fire in a timely and accurate manner whether there are diesel and other combustion aids at the fire scene. In this paper, polyethylene (PE) plastic products-plastic wrap and polyethylene raw materials were used as research objects to explore the influence of them on the analysis and identification of diesel fuel, an ignition booster at the scene of arson. Solid phase microextraction-gas chromatography mass spectrometry (SPME-GC-MS) was used to compare and analyze the characteristic spectra of samples and diesel fuel in three different combustion states: bulk body, incomplete combustion, complete combustion to spontaneous combustion extinguishing. Through the comparative analysis of polyethylene raw materials and PE plastic wrap body, it is found that the main components contained in the two components are basically the same, and the individual components contained in PE plastic wrap are not detected in polyethylene raw materials, mainly due to the addition of some additives in the production process of PE plastic wrap. Incomplete combustion of plastic wrap in the three combustion states has the greatest effect on diesel. When m/z=85, 178, 192, 183, the two contain the same components and the distribution law is basically the same, but when the ion m/z=83 is extracted, the sample of incomplete combustion plastic wrap does not contain the same naphthene as diesel. The application of SPME-GC-MS technology can distinguish incomplete combustion plastic wrap from diesel, which is convenient and fast, has high sensitivity and fast analysis speed, and can be used as a preferred method for experimental preliminary screening, which is suitable for the analysis of combustion promoters in the fire field.

With the rapid development of science and technology and the Internet, digital images are playing a more indispensable role in our life. Digital images are increasingly inseparable from social media, forensics, journalism, scientific research, medicine, and all sorts of industries. However, due to the high editability of digital image, anyone can modify the content of digital image arbitrarily through the image editing program, which makes the authenticity and objectivity of the image become unknown. Therefore, how to identify the authenticity of the image has become a hot topic among experts and scholars. In the recognition of image authenticity, the tampering form of modifying the image content by copying and pasting the same image is very common in many cases of image tampering. Aiming at such problems, this paper attempts to identify the copy-pasted tampering image of the same image based on LBP algorithm. Firstly, the corner points in the tampered image were identified by Harris corner detection technology, and a rectangular window with corner points as the core was constructed. Then, the rectangular window features were obtained by local binary pattern (LBP) technology, and the feature vectors were determined. Then, the feature vectors were sorted lexicographically. The image tampered area was determined by the similarity of adjacent feature vectors, and the image tampered area was determined by the similarity of adjacent feature vectors. Finally, wrong mismatching points are removed by RANSAC operator. Experimental results demonstrate that the algorithm can effectively check the degree of image tampering, and has high robustness for images tampered with by complex behaviors such as rotation tampering, brightness tampering, fuzzy tampering and noise tampering. Although the copy-paste source of multi-region tampering cannot be determined, the multi-region tampering region can be identified. The method can verify the same image copy-paste tampering image effectively.

Based on gas chromatography mass spectrometry and multivariate analysis, non-targeted profiling of methamphetamine in China was established to deduce the synthetic routes. After optimizing analysis condition with gas chromatography mass spectrometry, following appropriate examination of all the peaks, 32 impurities were selected as the specific components in seized methamphetamine samples. Unsupervised (principal component analysis, PCA; hierarchical cluster analysis, HCA) and supervised (orthogonal projections to latent structures-DA, OPLS-DA) techniques were employed for classifying the 190 MA seizures. The results of PCA, HCA and OPLS-DA were in good agreement and showed a good tool for analyzing seizures correlation and synthesis routes. The established non-targeted screening technology has the advantage of simple operation, high reliability, not needing standard materials of impurities and can provide informative support to crack down the methamphetamine-related crime.

There are few morphological methods for fingerprint age analysis, and it is not certain that the mastoid line morphology of fingerprint is correlated with age. Through statistical and correlated analysis of quantifiable feature information, it is found that the number of wrinkles and fine line, the density of mastoid line, the width of mastoid line and small furrow, and other indicators are correlated with age, but the use of these indicators is not enough to analyze the age of the hand imprint remnant. The purpose of this paper is to study the performance characteristics of fingerprints in different age groups, especially the age-related characteristic information that is difficult to be quantified in fingermarks at the present stage, such as the fuzzy phenomenon of fingermarks, the symmetry degree of the edge of friction ridges, and the change rule of sweat holes with age. We explore the specific method of age analysis by fingerprints through observing and summarizing. The fingerprints were observed through the microscope, and the typical examples were visually displayed by fingerprint atlas. Fingerprints are extracted from a large number of samples and represent the characteristics of most samples of the corresponding age. In this paper, the characteristics and performance of different age groups are introduced through text, and the characteristics of fingerprinting performance in different age groups are displayed more intuitively and vividly with the fingerprint atlas. After observing a large number of fingerprint samples in different age groups in the fingerprint archives, the morphological characteristics of fingerprint in different age groups were summarized, and the specific steps and methods of age analysis are preliminarily formed. The specific process of such analysis was showed in combination with practical cases. The map of fingerprint contains the various typical and atypical manifestations of the fingermark samples at the corresponding age. Detailed steps are proposed for a morphological method of age analysis using handprints, which, like a code of practice, can be used by an experienced fingerprint analyst to make a preliminary estimate of the age of the person left behind. This paper provides a direction for researching fingerprint-based age analysis.

With the breakthrough of deep-learning technology in the field of artificial intelligence, deepfake portrait videos appear more and more frequently, such as facial tampering, pornographic video face swapping, changing politicians’ faces and making false statements, etc. This kind of deepfakes may pose a threat to societies; therefore, distinguishing deepfake videos from genuine ones has become an urgen issue. Lots of deepfake detection methods are carried out by constructing many data sets with different compression factors. At present, the deepfake detection technology based on deep-learning algorithm is popular, which requires lots of significant time consumption and massive computing power for training classification model. At the same time, the attributes of black box and unexplainability of deep learning networks also plague the researchers in forensic science. In order to solve the problem of authenticity forensics of the deepfake portrait videos, this paper takes portrait videos encoded by H.264/AVC as the research object, and proposes a method based on inter-frame quantization parameter intensity value to detect deepfake portrait videos and real portrait videos. The selection of inter-frame quantization parameter intensity value and the determination of the inter-frame quantization parameter intensity by binary Logistic regression equation are expounded in detail. The experimental results show favorable accuracy and robustness for the deepfake portrait videos synthesized by DeepFaceLab platform. The paper proposed an interpretable detection method for deepfake portrait videos, which is conducive to determine the direction of investigation and confirm the criminal facts. But there are some limitations. Firstly, the experimental samples are not rich enough. Secondly, the introduced method is greatly affected by video compression, which caused limited application scenarios. Thirdly, the analysis efficiency needs to be optimized.

To establish an accelerated solvent extraction liquid chromatography-tandem mass spectrometry (ASE-HPLC-MS/MS) method for detecting MDMB-4en-PINACA and its metabolites in biological samples, the effects of liquid-liquid extraction and accelerated solvent extraction on drugs in liver and kidney tissue were investigated. Then the MDMB-4en-PINACA and its metabolites in liver and kidney tissue were analyzed by HPLC-MS/MS. The results showed the ASE had the best performance with a good linear relationship of the MDMB-4en-PINACA, and its hydrolytic metabolite (M1) and dealkenyl metabolite (M2) with the peak area (r > 0.997 3) in liver and kidney tissue within the range of 0.5-200 ng/g. The LOD values of MDMB-4en-PINACA, M1 and M2 were all 0.01 ng/g in liver tissue. The lower limit of detection values of MDMB-4en-PINACA, M1 and M2 were 0.01, 0.03 and 0.02 ng/g in kidney tissue, respectively. The method has the advantages of easy operation and high recovery. It could be concluded that the ASE-HPLC-MS/MS can be used to detect the MBMD-4en-PINACA and its metabolites in biological samples.

This study is aimed to fully explore the “time scale” role of hair samples in forensic science, provide technical methods for fine forensic analysis of hair samples, and offer reference examples for the description of drug history and criminal investigation of estazolam, a benzodiazepine antipsychotic. Actual hair samples were collected from adult estazolam users for time-sharing and sectional quantitative examination. Hairs were fixed and divided at the millimetre level using double-sided tape and a paper cutter. The hair-grinding extracts were detected by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The content of estazolam was determined by the internal standard curve method with diazepam-d5 as the internal standard. The limit of quantitation (LOQ) was 0.4 pg/mg for the hair segments above 1 cm and 2 pg/mg for the millimetre scale ones, respectively. The content distribution of estazolam along the hair shaft was investigated under long-term use, short-term use and drug withdrawal conditions. After a single dose of estazolam, continuous shaving sampling and analysis were carried out on 2 volunteers to monitor the dynamic process of estazolam in the hair growing out of the skin. In the hair samples collected in this study, after long-term treatment (more than 2 years), the highest content of estazolam was located in the near root or middle segments, and there was a general loss in the end segments. Hair dyeing and perm can change the hair matrix, leading to an accelerated loss from estazolam out of the hair. Under a single dose and short-term administration, there was a corresponding relationship between the concentrated distribution zone of estazolam and the time of administration, which shifted distally with the hair growth. The peak spacing could be used to estimate the hair growth rate of the subjects. After drug withdrawal, the content of estazolam decreased slowly from the hair tip to the hair root, and the elimination of estazolam in the proximal hair segment took more than 3 months. Shaved moustache samples may be deeply contaminated by self-sebum and sweat, resulting in the highest content occurring on the next day of drug administration. In different subjects, there was a positive correlation between the peak content of estazolam and the dose in the corresponding period. The peak contents of estazolam were generally higher than 60 pg/mg under long-term medication (above 7 mg per week) and lower than 20 pg/mg under single dose (1 mg or less) and short-term medication. The time-sharing and segmented test of hair can help to explore the time-space distribution of estazolam after it enters the body and provides theoretical and technical support for donor behavior association and drug history description.

In recent years, with the rapid development of artificial intelligence, the intelligent recognition of barefoot footprints and sock footprints has achieved good results. However, these two types of footprint recognition methods need to take off shoes, and the application scenarios are greatly limited. Single shoe-wearing footprint recognition due to the wide variety of shoe sole patterns and the random changes of patterns that cause a great obstacle to intelligent recognition, and the recognition accuracy rate is generally low. Shoe-wearing footprint recognition has become a challenging task. To solve this issue, we focus on the footprint recognition of different people wearing the same kind of shoes and propose a shoe-wearing footprint recognition network based on multi-scale feature fusion. This paper focuses on the problem of shoe-wearing footprint recognition; we collect a large number of footprint samples and create a shoe-wearing footprint dataset by rotating, panning, and adding noise to simulate the possible changes in the footprint images of the crime scene. Then we use ResNet as the backbone network and fully fuse the deep and shallow features of the footprint images by using the bidirectional pyramid feature fusion module. Finally, to solve the problem of reduced recognition accuracy of the network due to the variation of shoe-wearing footprint patterns, we propose a transfer learning method to quickly fit a model, which allows the network to learn the relationship between unknown footprint patterns and existing footprint patterns. It froze most of the convolutional layers of the pre-trained model and trained only a small number of convolutional and fully connected layers. After training tests, the shoe-wearing footprint recognition network based on multi-scale feature fusion achieves 93.1% recognition accuracy on the homemade footprint dataset, and the CMC evaluation index is also significantly better than other networks. When faced with footprint patterns for which the network model has not been trained, the method of transfer learning has higher recognition accuracy and faster speed than retraining such footprint patterns. Through extensive experiments, the shoe-wearing footprint recognition network based on multi-scale feature fusion has achieved good results. When faced with untrained footprint patterns, the transfer learning method has higher recognition accuracy and faster speed than retraining by collecting a large number of footprint samples of different patterns. Of course, the more new footprint pattern samples are used for transfer learning, the higher the accuracy of the final model will be.

Freeze drying technology can remove about 99% moisture from the substance, resulting in the freeze-dried product with an increased internal surface area, which is loose, porous, and easy to recover. These characteristics are very suitable for application in the PCR amplification kit of forensic STR examination. Lyophilized reagent could break the challenge of the upper limit of template addition in conventional PCR systems. This study aims to test the technical performance of DNATyper^TM 30 lyophilized reagent and evaluate its application ability in practical forensic cases. In order to test the detection rate, sensitivity, stability and other aspects of lyophilized reagents, an exploration was carried out by setting different storage conditions, time ranges and concentration gradient. In order to meet the needs of room temperature transportation, the reagents were placed at room temperature (23 °C to 25 °C) to simulate the transportation environment and compare the STR typing results at 1, 7, 15, 30 days. In sensitivity test, standard DNA 9947A was used as template and its gradient was set from 0.0313 ng to 2ng. Furthermore, three types of case samples were selected, including trace DNA sample, to test the practical application of lyophilized reagents and liquid reagents were used in parallel for comparison. The results showed that DNATyper^TM 30 lyophilized reagent had accurate typing results and the peak value was between 3000 and 20000 RFUs with good repeatability. The lyophilized reagent was validated to get a well-balanced STR typing result, and the sensitivity was 0.125 ng, while it showed good stability when stored at room temperature for over 1 month. Essentially, lyophilized reagents performed outstandingly in the detection of trace DNA samples, with a higher detection rate and higher peak height of loci compared to conventional reagents. Besides, compared to longer segment loci, lyophilized reagents had a higher detection rate for shorter segment loci. Overall, DNATyper^TM 30 lyophilized reagent, which is easy to operate and has a high sensitivity, can significantly increase the loading quantity of DNA tamplate, and extend the storage time at room temperature, and meet the needs of DNA tracing testing, long-distance transportation and room temperature carrying at the same time. Therefore, lyophilized reagents have broad prospects in forensic examination and good practical application value.

Recordings made by mobile phone are a common form of evidence in current litigation, but it is easy to separate recording files from their initial equipment, so the evidence review of originality and integrity of recordings is prone to cause disputes, which often need to be resolved through judicial expertise. However, the current achievements mostly focus on the integrity identification of mobile phone recordings and the equipment appraisal of IOS recording files. For the research on the traceability analysis of Android recording files, there is still a lack of research results. Because of this, we selected several Huawei and Xiaomi smartphones, simulated the typical operation behaviors such as recording, deleting, collecting, and transferring ordinary recordings and call recordings, and proposed a method for recording equipment identification & state analysis based on the operating system files, application logs and specific backup data through experiments. The experimental results show data recording methods and contents differences between Huawei and Xiaomi smartphones. However, under the recording process, the two types of mobile phones will write data records related to the recording files in specific operating system files. Therefore, we can identify the original recording equipment of specific mobile phone recording files and their changes by analysing system files, application logs, and even the particular backup data. This method is effective for recording devices’ identification of ordinary and call recordings made by Huawei and Xiaomi Android smartphones and their state analysis; Furthermore, it can also be used for equipment identification & state analysis for Harmony OS recordings and other multimedia files in Android smartphones. Of course, since many mobile phone manufacturers have made secondary development of the Android operating system, we must continue research equipment identification & state analysis of recordings made by other Android smartphone brands.

Writing robot is a kind of writing equipment; it came into being with the development of electronic engraving technology, 3D printing technology and computer simulation technology. According to their different structures, they can be divided into three types: straight joint writing robot, bending joint writing robot, planar joint writing robot. In recent years, due to the low price and good writing effect of the writing robot, this device has become popular, and may be used by some people to forge handwriting. Its potential risks have not only attracted the attention of domestic handwriting examiners, but also attracted the attention of some foreign handwriting examiners. For this problem, they conducted a lot of experimental research and theoretical discussion from the perspective of graphology. Based on the published papers at home and abroad, this paper first summarizes the types of writing robots and their respective working principles, and then summarizes the characteristics of the handwriting written by straight joint writing robots and bending joint writing robot. On this basis, four methods for recognizing robot handwriting are introduced, they are: overall feature recognition, detailed feature observation, analysis of handwriting overlap comparison, and computer automatic recognition. At the same time, this paper summarizes the social evaluation of robot’s ability to imitate human’s handwriting. Finally, it points out some problems existing in the existing research, which are mainly manifested in two aspects: first, there is a lack of in-depth analysis and research on the influencing factors of handwriting characteristics of writing robots, especially the lack of correlation analysis on how various writing parameters affect handwriting characteristics; second, there is a lack of necessary tracking research on some new writing robots. In view of these problems, the future research trend in this field is prospected, and the development trend of this equipment should be closely watched to prevent potential risks.

The paper aims to provide the reference and pathological data for the determination of the poisoning caused by heartbreak grass. Post-mortem examination, histopathological examination, and toxicological examination were performed on a deceased case of gelsemium elegans poisoning caused by ingestion of homemade Chinese medicine, and the tissue cells were stained and observed with microscope for pathological examination, while liquid chromatography mass spectrometry (LC-MS) was used for toxicological examination of heart blood and the residue of Chinese medicine. The relevant literature on gelsemium poisoning was reviewed，such as the poisoning symptoms，pathological feature and postmortem content of gelsemium in blood or tissue. Although many poisoning cases were reported, most of them were published by medical and health departments, and post-mortem toxicological data are scarce. This paper reported a dead case which gelsemine, koumine and gelsecine were detected in the heart blood with the concentration of 77.4, 30.3 and 37.1 ng/mL respectively, and the pathological examination showed that the SA node artery stenosis and myocardial fibrosis were found, but on typical toxic characteristics in tissues were observed. In this case of acute poisoning of Gelsemium elegans, no typical features were found by autopsy, but the obtained basic data such as toxin concentration in blood contributed to accumulate forensic toxicological data of Gelsemium elegans poisoning, and provided a reference for forensic toxicologists in identifying the cause of death in related cases.

This study is intended to provide a visual and quantifiable auxiliary technology for analysis of rib fracture injury mode. The rib fracture caused by boxing and chest compression is simulated by the finite element method: (1) The fist model is set to hit the left chest of THUMS4.02 (Total Human Model for Safety 4.02) at the speed of 6m/s. (2) The pressing hammer model is set to presses the precordial region of THUMS4.02 with a vertical downward force of 400 N. Distribution and morphology of rib fractures shown by the simulation results were compared, and the injury mechanism of the two was analyzed from the perspective of biomechanics. The simulation results are compared with those reported in the literature. The simulation results show that: (1) Shape and distribution characteristics of rib fractures: the shape of the broken ends of rib fractures caused by boxing is inward bending, and the fractures are distributed at the direct hitting positions of the ribs. The shape of the broken ends of rib fractures caused by chest compression is outward bending, and the fractures are distributed in the nondirect stress areas of the ribs. (2) Biomechanical response characteristics: when boxing the chest, the strain is concentrated in the position of the boxing ribs. Compressive stress is generated on the outer side of the rib at the striking position, and tensile stress is generated on the inner side. When pressing the chest, in addition to the strain distribution at the pressing position, there is another obvious strain distribution at the front line of the rib axillary on the pressing side of the chest. The lateral surface of the rib at the axillary front line generates tensile stress, and the medial surface causes compressive stress. The literature supports the simulation results, suggesting that the finite element method can be used to realize visual and quantifiable auxiliary analysis and interpretation of the injury mode of rib fracture.

The fingerprint analysis of age is one of the important research directions in the field of forensic science, but there are some problems with various existing analysis methods, such as strong subjective experience, low feature utilization, small sample size and poor classification effect. To solve the above problems, an age classification model based on random forest algorithm was proposed. Firstly, the right-hand fingerprint samples of 1 600 males (aged 11 to 76) were collected to quantitatively process the feature data that might change with age. Then, the correlation analysis between each feature’s data and individual age was carried out to screen for features that were significantly correlated with age. Finally, based on random forest, ExtraTrees, CatBoost, KNN and BP neural network algorithms, four and six age classification models were established respectively. The applicability of the random forest algorithm was evaluated by model comparison, and the importance ranking of characteristic variables was output. The experimental results show that the random forest algorithm is the best for classification among the five classification models, which are as follows: in the four-classification model, the classification accuracy of random forest is 0.812, and the Kappa coefficient is 0.738; in the six-classification model, the classification accuracy of random forest is 0.738, and the Kappa coefficient is 0.675. Therefore, the fingerprint age classification model based on the random forest algorithm proposed in this paper can judge the age range of individuals more scientifically and accurately, so as to provide clues and ranges for investigation and technical inspection.