A novel method of automated skull registration for forensic facial approximation

Modern forensic facial reconstruction techniques are based on an understanding of skeletal variation and tissue depths. These techniques rely upon a skilled practitioner interpreting limited data. To (i) increase the amount of data available and (ii) lessen the subjective interpretation, we use medical imaging and statistical techniques. We introduce a software tool, reality enhancement/facial approximation by computational estimation (RE/FACE) for computer-based forensic facial reconstruction. The tool applies innovative computer-based techniques to a database of human head computed tomography (CT) scans in order to derive a statistical approximation of the soft tissue structure of a questioned skull. A core component of this tool is an algorithm for removing the variation in facial structure due to skeletal variation. This method uses models derived from the CT scans and does not require manual measurement or placement of landmarks. It does not require tissue-depth tables, can be tailored to specific racial categories by adding CT scans, and removes much of the subjectivity of manual reconstructions.     

Validation of STR typing by capillary electrophoresis

With the use of capillary electrophoresis (CE), high-resolution electrophoretic separation of short tandem repeat (STR) loci can be achieved in a semiautomated fashion. Laser-induced detection of fluorescently labeled PCR products and multicolor analysis enable the rapid generation of multilocus DNA profiles. In this study, conditions for typing PCR-amplified STR loci by capillary electrophoresis were investigated using the ABI Prism 310 Genetic Analyzer (Applied Biosystems). An internal size standard was used with each run to effectively normalize mobility differences among injections. Alleles were designated by comparison to allelic ladders that were run with each sample set. Multiple runs of allelic ladders and of amplified samples demonstrate that allele sizes were reproducible, with standard deviations typically less than 0.12 bases for fragments up to 317 bases in length (largest allele analyzed) separated in a 47 cm capillary. Therefore, 99.7% of all alleles that are the same length should fall within the measurement error window of +/- 0.36 bases. Microvariants of the tetranucleotide repeats were also accurately typed by the analytical software. Alleles differing in size by one base could be resolved in two-donor DNA mixtures in which the minor component comprised > or = 5% of the total DNA. Furthermore, the quantitative data format (i.e., peak amplitude) can in some instances assist in determining individual STR profiles in mixed samples. DNA samples from previously typed cases (typed for RFLP, AmpliType PM+DQA1, and/or D1S80) were amplified using AmpFlSTR Profiler Plus and COfiler and were evaluated using the ABI Prism 310. Most samples yielded typable results. Compared with previously determined results for other loci, there were no discrepancies as to the inclusion or exclusion of suspects or victims. CE thus provides efficient separation, resolution, sensitivity and precision, and the analytical software provides reliable genotyping of STR loci. The analytical conditions described are suitable for typing samples such as reference and evidentiary samples from forensic casework.     

Non-fatal injuries due to interpersonal violence in Johannesburg-Soweto: incidence, determinants and consequences.

A prospective, hospital based sampling procedure conducted between June 8th, 1989 and August 24th, 1990 identified 1592 victims of interpersonal violence in six state and five private hospitals serving the Johannesburg magisterial district (which includes Soweto). For the 1282 victims resident in the area, this corresponded to crude annualised incidence rates per 100,000 population of 3821 for coloureds, 1527 for blacks, 467 for whites, 433 for asians and 1380 for all residents. Males, most of whom were aged 20-24 years, constituted 83.9% of all victims and were most often attacked on the streets. The majority of females were attacked at home by a spouse or lover, and most incidents occurred between dusk and midnight on weekends. Sharp violence was the most frequent type, involving 52.2% of the male and 51.4% of the female victims. Analysis by racial classification of the overall findings for mechanism of injury and relationship to aggressor showed dramatic differences. Limitations of the findings are discussed. After comparison with similar data for two United States cities and Copenhagen, Denmark, it is hypothesised that the racial differences in the present findings reflect the interplay of universal structural determinants and a specific colonial factor rooted in the sociology and psychology of oppression. Some implications of these findings and hypothesised determinants for violence prevention initiatives are mentioned.     

Political Studies and Cultural Studies: A Response to Finlayson and Martin

Cultural Studies has more to learn from Political Studies than Political Studies had to learn from Cultural Studies. In a recent issue of Politics (Vol 17, No 3), Finlayson and Martin argued that Political Studies should learn from the conceptions of 'culture' and 'politics' employed in Cultural Studies. This response reviews the development of Cultural Studies in order to elucidate its approach and argues that alternative approaches to the issue of culture are more useful for students of politics than Cultural Studies approaches.