Accuracy and reproducibility of forensic tire examination decisions.

Tire impression evidence can be a valuable tool during a crime scene investigation-it can link vehicles to scenes or secondary locations, and reveal information about the series of events surrounding a crime. The interpretation of tire impression evidence relies on the expertise of forensic tire examiners. To date, there have not been any studies published that empirically evaluate the accuracy and reproducibility of decisions made by tire impression examiners. This paper presents the results of a study in which 17 tire impression examiners and trainees conducted 238 comparisons on 77 distinct questioned impression-known tire comparison sets (QKsets). This study was conducted digitally and addressed examinations based solely upon the characteristics of the tire impression images provided. The quality and characteristics of the impressions were selected to be broadly representative of those encountered in casework. Participants reported their decisions using a multi-level conclusion scale: 68% of responses were class associations (Association of Class Characteristics or Limited Association of Class), 21% were definitive decisions (ID or Exclusion), 8% were probable decisions (High Degree of Association or Indications of Non-Association), and 3% were neutral responses (Not Suitable or Inconclusive). Although class associations were the most reported response type, when definitive decisions were reported, they were often correct: 96% of IDs and 89% of Exclusions were consistent with ground truth regarding the source of the known tire in the QKset. Overall, we observed 4 erroneous definitive decisions (3 Exclusions on mated QKsets; 1 ID on a nonmated QKset) and 1 incorrect probable decision (Indications of Non-Association on a mated QKset). Decision rates were notably associated with both quality (lower quality questioned impressions were more likely to result in class associations) and dimensionality (2D questioned impressions were more likely to result in definitive decisions), which were correlated factors. Although the study size limits the precision of the measured rates, the results of this study remain valuable to the forensic science and legal communities and provide empirical data regarding examiner performance for a discipline that previously did not have any such estimates.

Recognizability of Demographically Altered Computerized Facial Approximations in an Automated Facial Recognition Context for Potential Application in Unidentified Persons Data Repositories.

This study examined the recognizability of demographically altered facial approximations for potential utility in unidentified persons tracking systems. Five computer-generated approximations were generated for each of 26 African male participants using the following demographic parameters: (i) African male (true demographics), (ii) African female, (iii) Caucasian male, (iv) Asian male, and (v) Hispanic male. Overall, 62% of the true demographic facial approximations for the 26 African male participants examined were matched to a corresponding life photo within the top 50 images of a candidate list generated from an automated blind search of an optimally standardized gallery of 6159 photographs. When the African male participants were processed as African females, the identification rate was 50%. In contrast, less congruent identification rates were observed when the African male participants were processed as Caucasian (42%), Asian (35%), and Hispanic (27%) males. The observed results suggest that approximations generated using the opposite sex may be operationally informative if sex is unknown. The performance of approximations generated using alternative ancestry assignments, however, was less congruent with the performance of the true demographic approximation (African male) and may not yield as operationally constructive data as sex-altered approximations.

Accuracy and reliability of forensic handwriting comparisons.

Forensic handwriting examination involves the comparison of writing samples by forensic document examiners (FDEs) to determine whether or not they were written by the same person. Here we report the results of a large-scale study conducted to assess the accuracy and reliability of handwriting comparison conclusions. Eighty-six practicing FDEs each conducted up to 100 handwriting comparisons, resulting in 7,196 conclusions on 180 distinct comparison sets, using a five-level conclusion scale. Erroneous "written by" conclusions (false positives) were reached in 3.1% of the nonmated comparisons, while 1.1% of the mated comparisons yielded erroneous "not written by" conclusions (false negatives). False positive rates were markedly higher for nonmated samples written by twins (8.7%) compared to nontwins (2.5%). Notable associations between training and performance were observed: FDEs with less than 2 y of formal training generally had higher error rates, but they also had higher true positive and true negative rates because they tended to provide more definitive conclusions; FDEs with at least 2 y of formal training were less likely to make definitive conclusions, but those definitive conclusions they made were more likely to be correct (higher positive predictive and negative predictive values). We did not observe any association between writing style (cursive vs. printing) and rates of errors or incorrect conclusions. This report also provides details on the repeatability and reproducibility of conclusions, and reports how conclusions are affected by the quantity of writing and the similarity of content.

Cranial and facial inter-landmark distances and tissue depth dataset from computed tomography scans of 388 living persons.

Computed tomography (CT) scans of 388 living adults of both sexes were collected from four self-identified ancestry groups from the United States (African, Asian, European, and Hispanic). Scans were acquired from multiple institutions and under a variety of scanning protocols. Scans were used to produce 3D bone and soft tissue models, from which were derived cranial and facial inter-landmark distances (ILDs) and soft tissue depth measurements. Similar measurements were made on 3D facial approximations produced by ReFace software. 3D models and all measurements were obtained using MimicsR software. These measurements are useful for facial approximations of unidentified decedents and for investigations into human variation between and among ancestry groups and sexes.

Accuracy and reproducibility of conclusions by forensic bloodstain pattern analysts.

Although the analysis of bloodstain pattern evidence left at crime scenes relies on the expert opinions of bloodstain pattern analysts, the accuracy and reproducibility of these conclusions have never been rigorously evaluated at a large scale. We investigated conclusions made by 75 practicing bloodstain pattern analysts on 192 bloodstain patterns selected to be broadly representative of operational casework, resulting in 33,005 responses to prompts and 1760 short text responses. Our results show that conclusions were often erroneous and often contradicted other analysts. On samples with known causes, 11.2% of responses were erroneous. The results show limited reproducibility of conclusions: 7.8% of responses contradicted other analysts. The disagreements with respect to the meaning and usage of BPA terminology and classifications suggest a need for improved standards. Both semantic differences and contradictory interpretations contributed to errors and disagreements, which could have serious implications if they occurred in casework.

Recognizability of computer-generated facial approximations in an automated facial recognition context for potential use in unidentified persons data repositories: Optimally and operationally modeled conditions.

Currently in the United States, the remains of thousands of unidentified human decedents are housed in medical, law enforcement, and forensic facilities throughout the country. A number of digital data repositories have been established to curate and disseminate the details of these unidentified decedent cases; some repositories also maintain records of missing persons. Although a cross-reference for textual data similarity occurs between the missing persons and unidentified decedent records in some repositories, no repository is currently known to employ an image analysis technology for cross-referencing image data. Results suggest that the computer-generated facial approximations used in this research were consistently included in prioritized candidate lists when used in an automated facial recognition context. Two concurrent studies exploring the specific use-case discussed here were executed. The first employed an optimally-conditioned facial image gallery (g=6159) (i.e., a gallery comprised of highly consistent facial images), a research design intended to establish the ceiling performance of the combined use of the two software programs employed. The second employed a gallery (g=1816) compiled from a real-world dataset of missing persons' facial images, a research design intended to inform potential operational performance when using the highly varied facial images typically comprising public databases. Multiple types of facial approximations (reconstructions) with varying degrees of weight adjustments, age adjustments, or the presence (or absence) of visible eyes, and combinations of these variables, were evaluated. Overall, in the larger, optimally modeled study, 53% of the facial approximations for the t=159 test subjects examined were matched to his or her corresponding life photo within the top 50 images of a candidate list generated from a blind (unrestricted) search of the highly consistent gallery (g=6159). In the operationally modeled study, 31% of the test subjects' (t=16) facial approximations were matched to their corresponding life photos within the top 50 images of a candidate list generated from a blind search of the gallery populated with images from an operational dataset (g=1816). As anticipated, candidate list inclusion rates improved with the use of demographic filters. No significantly different inclusion rates were observed between the sex or age cohorts examined. Significant differences were, however, observed across population cohorts. Entities curating missing and unidentified decedent records may benefit from a paired implementation of facial recognition technology and computer-generated approximations as part of a comprehensive investigative strategy for the specific envisioned use-case discussed in this research.

Biometric correspondence between reface computerized facial approximations and CT-derived ground truth skin surface models objectively examined using an automated facial recognition system.

This study employed an automated facial recognition system as a means of objectively evaluating biometric correspondence between a ReFace facial approximation and the computed tomography (CT) derived ground truth skin surface of the same individual. High rates of biometric correspondence were observed, irrespective of rank class (Rk) or demographic cohort examined. Overall, 48% of the test subjects' ReFace approximation probes (n=96) were matched to his or her corresponding ground truth skin surface image at R1, a rank indicating a high degree of biometric correspondence and a potential positive identification. Identification rates improved with each successively broader rank class (R10=85%, R25=96%, and R50=99%), with 100% identification by R57. A sharp increase (39% mean increase) in identification rates was observed between R1 and R10 across most rank classes and demographic cohorts. In contrast, significantly lower (p<0.01) increases in identification rates were observed between R10 and R25 (8% mean increase) and R25 and R50 (3% mean increase). No significant (p>0.05) performance differences were observed across demographic cohorts or CT scan protocols. Performance measures observed in this research suggest that ReFace approximations are biometrically similar to the actual faces of the approximated individuals and, therefore, may have potential operational utility in contexts in which computerized approximations are utilized as probes in automated facial recognition systems.

Automated facial recognition of manually generated clay facial approximations: Potential application in unidentified persons data repositories.

This research examined how accurately 2D images (i.e., photographs) of 3D clay facial approximations were matched to corresponding photographs of the approximated individuals using an objective automated facial recognition system. Irrespective of search filter (i.e., blind, sex, or ancestry) or rank class (R1, R10, R25, and R50) employed, few operationally informative results were observed. In only a single instance of 48 potential match opportunities was a clay approximation matched to a corresponding life photograph within the top 50 images (R50) of a candidate list, even with relatively small gallery sizes created from the application of search filters (e.g., sex or ancestry search restrictions). Increasing the candidate lists to include the top 100 images (R100) resulted in only two additional instances of correct match. Although other untested variables (e.g., approximation method, 2D photographic process, and practitioner skill level) may have impacted the observed results, this study suggests that 2D images of manually generated clay approximations are not readily matched to life photos by automated facial recognition systems. Further investigation is necessary in order to identify the underlying cause(s), if any, of the poor recognition results observed in this study (e.g., potential inferior facial feature detection and extraction). Additional inquiry exploring prospective remedial measures (e.g., stronger feature differentiation) is also warranted, particularly given the prominent use of clay approximations in unidentified persons casework.

Application of Enhanced Point Estimators to a Sample of In Vivo CT-derived Facial Soft Tissue Thicknesses.

Facial approximations based on facial soft tissue depth measurement tables often utilize the arithmetic mean as a central tendency estimator. Stephan et al. (J Forensic Sci 2013;58:1439) suggest that the shorth and 75-shormax statistics are better suited to describe the central tendency of non-normal soft tissue depth data, while also accommodating normal distributions. The shorth, 75-shormax, arithmetic mean, and other central tendency estimators were evaluated using a CT-derived facial soft tissue depth dataset. Differences between arithmetic mean and shorth mean for the tissue depths examined ranged from 0 mm to +2.3 mm (average 0.6 mm). Differences between the arithmetic mean plus one standard deviation (to approximate the same data points covered by the 75-shormax) and 75-shormax values ranged from -0.8 mm to +0.7 mm (average 0.2 mm). The results of this research suggest that few practical differences exist across the central tendency point estimators for the evaluated soft tissue depth dataset.

Accuracy Rates of Ancestry Estimation by Forensic Anthropologists Using Identified Forensic Cases.

A common task in forensic anthropology involves the estimation of the ancestry of a decedent by comparing their skeletal morphology and measurements to skeletons of individuals from known geographic groups. However, the accuracy rates of ancestry estimation methods in actual forensic casework have rarely been studied. This article uses 99 forensic cases with identified skeletal remains to develop accuracy rates for ancestry estimations conducted by forensic anthropologists. The overall rate of correct ancestry estimation from these cases is 90.9%, which is comparable to most research-derived rates and those reported by individual practitioners. Statistical tests showed no significant difference in accuracy rates depending on examiner education level or on the estimated or identified ancestry. More recent cases showed a significantly higher accuracy rate. The incorporation of metric analyses into the ancestry estimate in these cases led to a higher accuracy rate.

Automated Facial Recognition of Computed Tomography-Derived Facial Images: Patient Privacy Implications.

The recognizability of facial images extracted from publically available medical scans raises patient privacy concerns. This study examined how accurately facial images extracted from computed tomography (CT) scans are objectively matched with corresponding photographs of the scanned individuals. The test subjects were 128 adult Americans ranging in age from 18 to 60 years, representing both sexes and three self-identified population (ancestral descent) groups (African, European, and Hispanic). Using facial recognition software, the 2D images of the extracted facial models were compared for matches against five differently sized photo galleries. Depending on the scanning protocol and gallery size, in 6-61 % of the cases, a correct life photo match for a CT-derived facial image was the top ranked image in the generated candidate lists, even when blind searching in excess of 100,000 images. In 31-91 % of the cases, a correct match was located within the top 50 images. Few significant differences (p > 0.05) in match rates were observed between the sexes or across the three age cohorts. Highly significant differences (p < 0.01) were, however, observed across the three ancestral cohorts and between the two CT scanning protocols. Results suggest that the probability of a match between a facial image extracted from a medical scan and a photograph of the individual is moderately high. The facial image data inherent in commonly employed medical imaging modalities may need to consider a potentially identifiable form of "comparable" facial imagery and protected as such under patient privacy legislation.

Automated facial recognition and candidate list rank change of computer generated facial approximations generated with multiple eye orb positions.

Expanding on research previously reported by the authors, this study further examines the recognizability of ReFace facial approximations generated with the following eye orb positions: (i) centrally within the bony eye socket, (ii) 1.0mm superior and 2.0mm lateral relative to center, and (iii) 1.0mm superior and 2.5mm lateral relative to center. Overall, 81% of the test subjects' approximation ranks improved with the use of either of the two supero-lateral eye orbs. Highly significant performance differences (p<0.01) were observed between the approximations with centrally positioned eye orbs (i) and approximations with the eye orbs placed in the supero-laterally positions (ii and iii). Noteworthy was the observation that in all cases when the best rank for an approximation was obtained with the eye orbs in position (iii), the second best rank was achieved with the eye orbs in position (ii). A similar pattern was also observed when the best rank was obtained with the eye orbs in position (ii), with 60% of the second best ranks observed in position (iii). It is argued, therefore, that an approximation constructed with the eye orbs placed in either of the two supero-lateral positions may be more effective and operationally informative than centrally positioned orbs.

Accuracy Rates of Sex Estimation by Forensic Anthropologists through Comparison with DNA Typing Results in Forensic Casework.

A common task in forensic anthropology involves the estimation of the biological sex of a decedent by exploiting the sexual dimorphism between males and females. Estimation methods are often based on analysis of skeletal collections of known sex and most include a research-based accuracy rate. However, the accuracy rates of sex estimation methods in actual forensic casework have rarely been studied. This article uses sex determinations based on DNA results from 360 forensic cases to develop accuracy rates for sex estimations conducted by forensic anthropologists. The overall rate of correct sex estimation from these cases is 94.7% with increasing accuracy rates as more skeletal material is available for analysis and as the education level and certification of the examiner increases. Nine of 19 incorrect assessments resulted from cases in which one skeletal element was available, suggesting that the use of an "undetermined" result may be more appropriate for these cases.

Assessment of presentation methods for ReFace computerized facial approximations.

Facial approximations (whether clay sculptures, sketches, or computer-generated) can be presented to the public in a variety of layouts, but there are currently no clear indicators as to what style of presentation is most effective at eliciting recognition. The primary purpose of this study is to determine which of five presentation methods produces the most favorable recognition results. A secondary goal of the research is to evaluate a new method for assessing the accuracy of facial approximations. Previous studies have evaluated facial approximation effectiveness using standards similar to studies of eyewitness identification in which a single, definitive choice must be made by the research participant. These criteria seem inappropriate given that facial approximation is strictly an investigative tool to help narrow the search for potential matching candidates in the process of identification. Results from the study showed a higher performance for methods utilizing more than one image of the approximation, but which specific method performed best varied among approximation subjects. Also, results for all five presentation methods showed that, when given the opportunity to select more than one approximation, participants were consistently better at identifying the correct approximation as one of a few possible matches to the missing person than they were at singling out the correct approximation. This suggests that facial approximations have perhaps been undervalued as investigative tools in previous research.

Accuracy of standard craniometric measurements using multiple data formats.

UNLABELLED: With continuing advancements in biomedical imaging technologies, anthropologists are increasingly making use of data derived from indirect measurement and analysis of skeletal material. To that end, the purpose of this study was to test the reliability of 26 standard craniometric measurements routinely utilized in forensic casework across several different imaging technologies. Measurements from five crania of known individuals were collected in duplicate by two anthropologists via computed tomography (CT) scans and three-dimensional (3D) laser scans of the known skulls. The laser scans were also used to create prototype models of the known skulls. These prototypes were, themselves, laser-scanned, and measurements were also collected from the prototypes and the laser scans of the prototypes. Measurement sets from each technology were then compared with one another using the previously collected osteometric measurements taken on the crania themselves as the ground truth. RESULT: indicate that, while the majority of measurements showed no significant differences across data formats, a handful were found to be problematic for particular technologies. For instance, measurements taken in a supero-inferior direction (e.g., BBH, OBH) from CT scans were prone to greater deviation from direct measurements of the cranium than other technologies, especially for CT scans taken at 5 mm thickness and increment. Also, several measurements defined by Type 1 landmarks, particularly those occurring at complicated or indistinct suture junctures (e.g., ASB, ZMB), were found to have high variance across all technologies while measurements based on Type 3 landmarks proved to be highly reproducible. This is contrary to measurements taken directly on crania, in which measures defined by Type 1 landmarks are typically the most reliable, likely attributable to diminished or totally obscured suture definition in the scan data. If medical imaging data are to be increasingly utilized in anthropological studies, it may be prudent to bear in mind that the reliability of measurements taken on an actual skull may not be the same as for measurements taken from medical scans.

Preliminary assessment of facial soft tissue thickness utilizing three-dimensional computed tomography models of living individuals.

Facial approximation is the technique of developing a representation of the face from the skull of an unknown individual. Facial approximation relies heavily on average craniofacial soft tissue depths. For more than a century, researchers have employed a broad array of tissue depth collection methodologies, a practice which has resulted in a lack of standardization in craniofacial soft tissue depth research. To combat such methodological inconsistencies, Stephan and Simpson 2008 [15] examined and synthesized a large number of previously published soft tissue depth studies. Their comprehensive meta-analysis produced a pooled dataset of averaged tissue depths and a simplified methodology, which the researchers suggest be utilized as a minimum standard protocol for future craniofacial soft tissue depth research. The authors of the present paper collected craniofacial soft tissue depths using three-dimensional models generated from computed tomography scans of living males and females of four self-identified ancestry groups from the United States ranging in age from 18 to 62 years. This paper assesses the differences between: (i) the pooled mean tissue depth values from the sample utilized in this paper and those published by Stephan 2012 [21] and (ii) the mean tissue depth values of two demographically similar subsets of the sample utilized in this paper and those published by Rhine and Moore 1984 [16]. Statistical test results indicate that the tissue depths collected from the sample evaluated in this paper are significantly and consistently larger than those published by Stephan 2012 [21]. Although a lack of published variance data by Rhine and Moore 1984 [16] precluded a direct statistical assessment, a substantive difference was also concluded. Further, the dataset presented in this study is representative of modern American adults and is, therefore, appropriate for use in constructing contemporary facial approximations.

Preliminary performance assessment of computer automated facial approximations using computed tomography scans of living individuals.

ReFace (Reality Enhancement Facial Approximation by Computational Estimation) is a computer-automated facial approximation application jointly developed by the Federal Bureau of Investigation and GE Global Research. The application derives a statistically based approximation of a face from a unidentified skull using a dataset of ~400 human head computer tomography (CT) scans of living adult American individuals from four ancestry groups: African, Asian, European and Hispanic (self-identified). To date only one unpublished subjective recognition study has been conducted using ReFace approximations. It indicated that approximations produced by ReFace were recognized above chance rates (10%). This preliminary study assesses: (i) the recognizability of five ReFace approximations; (ii) the recognizability of CT-derived skin surface replicas of the same individuals whose skulls were used to create the ReFace approximations; and (iii) the relationship between recognition performance and resemblance ratings of target individuals. All five skin surface replicas were recognized at rates statistically significant above chance (22-50%). Four of five ReFace approximations were recognized above chance (5-18%), although with statistical significance only at the higher rate. Such results suggest reconsideration of the usefulness of the type of output format utilized in this study, particularly in regard to facial approximations employed as a means of identifying unknown individuals.

A study of the human decomposition sequence in central Texas.

Decomposition studies utilizing nonhuman subjects as human analogues are well established, but fewer studies utilizing intact human remains exist. This study provides data from a controlled decomposition study involving human remains in Central Texas. A 63.5-kg unmodified cadaver was placed in an open-air site and observed over a 10-week period from April 11 through June 19, 2008. A wire enclosure restricted scavenger access. State of decomposition and environmental conditions were recorded daily for the first 36 days and then every 2 weeks thereafter. Results indicated a high degree of correlation with other decomposition studies originating in the southwestern United States, although slight deviations for the average duration of early events were noted. The data were also utilized to test a quantitative method for estimating the postmortem interval. Results indicated preliminary support for a quantitative approach. Additional research is encouraged to further establish the human decomposition data set for Central Texas.