Personal Identification by Comparative Radiography and Frontal Sinus

Dental identification by comparative radiography sits primarily in the domain of forensic odontology, but forensic anthropologists must understand its principles because the two disciplines intersect in most DVI (Disaster Victim Identification) operations and in single-case homicide identifications where both types of specialist work from the same skull. The basic principle is the same as for frontal sinus comparison: an antemortem dental radiograph taken for clinical purposes is compared, feature by feature, with a postmortem radiograph taken from the recovered dentition.

The antemortem material most commonly used is a full-mouth periapical series from a general dentist, an orthopantomogram (OPT or panoramic radiograph) that shows all teeth in a single image, or bitewing radiographs that show interproximal contact areas and the coronal portion of the root. Each image shows, in addition to the clinical information for which it was taken, a detailed silhouette of root morphology, pulp chamber and canal shape, periodontal bone level, the geometry of any existing restorations, and the pattern of any root resorption, periapical pathology or developmental anomalies.

The comparison proceeds landmark by landmark. Root length and curvature, the width of the root canal relative to the root, the morphology of the pulp horn, the outline of any amalgam, composite or crown restoration, the position and shape of any periapical radiolucency, and the pattern of interseptal bone loss between teeth are each assessed for consistency or inconsistency between the antemortem and postmortem images. In the Sushil Sharma case of 1995, the partial remains recovered from a tandoor oven in a New Delhi restaurant were identified in part through dental comparison with radiographs obtained from the dentist of the victim, Naina Sahni. The dental and skull examination conducted at the AIIMS Forensic Medicine department contributed to the conviction that followed.

A key difference from frontal sinus comparison is the effect of time: dental structures change with age (restorations accumulate, teeth are extracted, root resorption progresses), so the examiner must account for the interval between the antemortem film and the postmortem examination. A missing tooth shown as present in a ten-year-old antemortem film is consistent with extraction in the interval, not with inconsistency of identification.

In the United Kingdom, the Stephen Lawrence case (1993 murder, 2011 convictions after DNA evidence) involved a later exhumation in which dental and other comparative methods were among the approaches reviewed. In Australia, the Victorian Institute of Forensic Medicine (VIFM) routinely coordinates dental radiographic comparison as part of its DVI protocol, and the Bali bombing of 2002 (88 Australian victims) relied on dental records as the primary identification method for many of the recovered remains, alongside the DNA programme coordinated through the AFP.

The modern skeleton is increasingly populated by metallic and polymeric hardware installed during surgical procedures. Total hip replacements, total knee replacements, femoral nailing for fracture fixation, spinal fusion cages, pedicle screws, orthopaedic plates for facial reconstruction, cochlear implant housings, pacemaker and defibrillator cases, and vascular stents all remain identifiable in skeletal remains long after soft tissue has decomposed. For the forensic anthropologist, each of these devices is a potential identification route, because each is manufactured to regulatory standards that require its individual or lot-level traceability.

The regulatory framework differs by jurisdiction. In the United States, the Food and Drug Administration (FDA) Unique Device Identification (UDI) system, mandated under the Food and Drug Administration Safety and Innovation Act 2012 and phased in from 2014 onward, requires that every medical device distributed in the US carry a unique device identifier on its labelling and in the FDA's Global Unique Device Identification Database (GUDID). The UDI encodes the manufacturer, the device model, the lot or batch number, and often the individual unit serial number, sufficient to trace the device to the specific surgical case if hospital records are available. In the European Union, the Medical Device Regulation (EU MDR 2017/745) introduced a parallel UDI requirement phased in from 2021, with devices registered in the European Database on Medical Devices (EUDAMED). In India, the Medical Devices Rules 2017 under the Drugs and Cosmetics Act include a UDI notification framework, though full implementation is more variable than in the US and EU.

The identification workflow for surgical hardware begins at the scene or at the laboratory when the device is recovered from the remains. The examiner documents the device type, notes all visible markings on its surface (manufacturer name or logo, model code, lot or serial number engraved or laser-etched on the metal), and photographs the device under raking light and macro illumination. For highly corroded or bone-encrusted devices, X-ray fluorescence (XRF) spectroscopy can identify the alloy (titanium, cobalt-chromium, stainless steel) and confirm whether visible characters are genuine manufacturer markings. The recovered code is then queried against the manufacturer's database and, where applicable, the regulatory database, to obtain the device model, manufacture date, batch, and distribution chain.

The critical step is then a hospital records search. The device can be traced to the supply chain, but to reach the patient level, the examiner needs the hospital implant log, the surgical theatre record, or the patient's medical summary, which specifies which device, from which batch, was implanted in which patient on which date. Data protection legislation in most jurisdictions (HIPAA in the US, GDPR in the EU, the DPDP Act 2023 in India) creates a framework for lawful access to such records in the context of a death investigation: the patient is deceased and the purpose is identification, not treatment.

1. Device recovery and documentation
   Photograph all surfaces in situ. Record device location relative to the skeleton. Assess corrosion grade. Note all visible markings. Sketch or digitally scan the device.
2. Marking identification
   Under macro lens or digital microscope, record manufacturer logo, model number, lot number, and serial number. Use XRF if corrosion obscures marks; alloy composition confirms the device class.
3. Manufacturer database query
   Contact the manufacturer's regulatory affairs department with the device codes. Obtain the model specifications, manufacturing date range, and the list of hospitals in the distribution batch (if lot-level tracing applies).
4. FDA / EUDAMED / national registry query
   Cross-check the UDI against the relevant regulatory database to confirm the device's legal distribution record. Confirms the device was not counterfeit.
5. Hospital records request
   With a formal legal request (search warrant, court order, coroner's authority), obtain the implant logs from hospitals in the device's distribution region for the relevant date range. Match to patient records.
6. Identification conclusion
   If the hospital record names the patient who received the specific device lot or serial number and the biological profile and other evidence are consistent, a positive identification can be established. Report in the categorical framework.

The Stephen Lawrence investigation in England, though primarily a homicide case rather than an identification case, illustrated how forensic examination of physical traces on a body can contribute to long-delayed legal proceedings. More directly relevant are cases such as the identification of victims in the 2005 London bombings (7 July), where devices recovered from some victims assisted identification alongside DNA and dental methods, and the 2013 Uttarakhand flood disaster in India, where forensic teams encountered the challenge of identifying decomposed remains against incomplete medical records.

Beyond the frontal sinus, the skeleton contains dozens of radiographically visible features that are sufficiently individual to support identification when an antemortem radiograph is available. The principle is the same as for frontal sinus comparison: the feature must be stable over the relevant time interval, sufficiently variable between individuals that the probability of two people sharing the identical feature pattern is low, and visible on both the antemortem and postmortem image at comparable angles.

Vertebral osteophytes are bony outgrowths from the vertebral body margins, usually resulting from degenerative disc disease or spondylosis. Their size, position, shape, and the specific vertebral levels at which they occur are highly individual. An antemortem lumbar spine radiograph taken for clinical back pain assessment may show a distinctive triangular osteophyte at L3-L4 or a bridging osteophyte at two adjacent levels that is immediately recognisable in the postmortem skeletal comparison. The SWGANTH 2013 comparative radiography guidance specifically includes vertebral morphology as an acceptable comparison medium.

Sinus pneumatisation extends to the mastoid process, the sphenoid sinus, and the ethmoid labyrinth, each of which shows individual variation in extent, cell pattern, and the presence of septa between cells. Mastoid cell patterns visible on skull radiographs have been used for identification since the 1950s, and the method is sufficiently established that several reference series comparing antemortem and postmortem mastoid films have been published from European and North American forensic medicine departments.

The sella turcica, the bony saddle in the sphenoid bone that houses the pituitary gland, varies considerably in its depth, breadth, and the shape of its posterior clinoid processes. A lateral skull radiograph taken for any clinical neurological purpose shows the sella turcica in profile; this image, compared with a lateral postmortem skull radiograph, provides a comparison medium that is independent of the frontal sinus and can serve as corroboration or, in the absence of frontal sinus films, as the primary comparison.

Calvarial suture morphology, including the detailed interlocking pattern of the coronal, sagittal, and lambdoid sutures visible in antero-posterior and lateral skull radiographs, is highly individual and stable in adults. Combined with trabecular bone patterns within the frontal and parietal bones (the digital impressions of meningeal vessels, the pattern of arachnoid granulation pits), the calvarial radiograph contains more individuating information than is often appreciated.

Healed fractures represent a distinct category of individuating features. A healed rib fracture from a childhood injury, visible as a periosteal callus on an antemortem chest radiograph, can be matched to the postmortem skeletal finding with a high degree of certainty if the location, size, and angulation of the callus correspond. The same applies to healed long-bone fractures, old compression fractures of vertebral bodies, and healed facial fractures visible on skull films.

The categorical reporting language used in comparative radiographic identification is not a legacy of excessive caution; it reflects the genuine epistemology of the comparison method. A frontal sinus comparison that shows concordance across all six morphological features with no unexplained discrepancy supports a conclusion that the antemortem and postmortem images derive from the same individual. The examiner cannot, from radiographic comparison alone, assign a statistical likelihood ratio in the same quantitative sense that a DNA STR analysis can, because the reference database for frontal sinus morphology, while growing, does not yet allow the precise population frequency calculations that STR typing provides.

This means that the reporting language in comparative radiographic identification is categorical rather than probabilistic. The three categories used by SWGANTH, ABFA-credentialed practitioners, and INTERPOL DVI teams are: positive identification (all features examined are consistent, no feature is inconsistent, the overall pattern of concordance is of sufficient rarity that the examiner is satisfied beyond a reasonable standard of scientific doubt); exclusion (one or more features are clearly inconsistent in a way that cannot be explained by projection angle, image quality, interval change, or disease, such that the antemortem and postmortem images cannot derive from the same individual); and inconclusive (the image quality, degree of preservation, or available features do not permit a determination in either direction).

In the United Kingdom Crown Court, radiographic identification evidence by a suitably qualified practitioner is admissible expert opinion under the expert witness framework established by the Criminal Practice Directions and reinforced by the Law Commission's 2011 Report on Expert Evidence. The evidence does not require a specific ABFA credential for admissibility, but the examiner must demonstrate sufficient training, experience, and familiarity with the comparison methodology to satisfy the court that their opinion has a reliable scientific basis. The UK position is essentially consistent with the US Daubert standard: reliability of the underlying scientific method is the threshold, not formal credentialing.

In India, comparative radiographic identification evidence is submitted under the framework of Section 45 of the Indian Evidence Act 1872 (now Section 39 of the Bharatiya Sakshya Adhiniyam 2023), which governs the admissibility of expert opinion. The expert must be qualified in the relevant field; the comparison process and the basis for the conclusion must be explained to the court. Indian courts have accepted both frontal sinus and dental radiographic comparison evidence in cases where it was presented by forensic medicine specialists from AIIMS, state FSLs, or equivalent institutions.