---
title: "CT-based Exploratory Analysis of Bronze Age Urns: Virtual vs Physical Assessment of Cremated Remai"
id: "plos-one-23-virtuality-vs-reality-testing-a-virtual-approach-for-the-exploratory-analysis"
canonical_url: "https://medichelpline.com/clinical-feed/plos-one-23-virtuality-vs-reality-testing-a-virtual-approach-for-the-exploratory-analysis"
content_type: "clinical_feed_article"
specialty: "General"
source_name: "PLOS ONE (Medicine)"
source_url: "https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0358226"
published_at: "2026-09-22T14:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# CT-based Exploratory Analysis of Bronze Age Urns: Virtual vs Physical Assessment of Cremated Remai
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/plos-one-23-virtuality-vs-reality-testing-a-virtual-approach-for-the-exploratory-analysis
- **Specialty:** [General](https://medichelpline.com/clinical-feed/general.md)
- **Primary Source:** PLOS ONE (Medicine)
- **Source URL:** [Original Journal Publication](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0358226)
- **Published At:** 2026-09-22T14:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- This study evaluated a relatively simple virtual workflow using **CT** scans to extract qualitative and quantitative data from urned cremated remains from a Middle–Late Bronze Age necropolis in Vicofertile, Parma, Italy. The sample comprised 10 non-micro-excavated urns scanned and then physically micro-excavated for comparison. - The virtual protocol aimed to retrieve information on urn taphonomy, biological profile indicators (age class and sex), **Minimum Number of Individuals (MNI)**, thermal alterations to bones, and the spatial (vertical and horizontal) distribution of skeletal elements within vessels. - Scans could reliably indicate age classes, burial taphonomy, and heat-induced bone changes such as fractures; these observations were confirmed by subsequent physical analysis. - The MNI estimated from scans was 10 but increased by one after physical excavation, indicating CT underestimated individual counts in at least one case. Sex estimation from CT data was not possible with the image quality used. - Spatial distribution of remains was assessed by anatomical categories (cranial, long bone, spongy bone, upper limb, lower limb, trunk). Vertical cranial distribution was the most accurately predicted from CT images; other categories were less well resolved. - Major limitations identified include overlapping grey values due to similar densities of materials within urns, scan artifacts, and overall low-to-average image quality that complicates segmentation and material discrimination. - The authors present a segmentation strategy intended to be relatively simple and faster than more complex semi-automated methods; however, manual corrections and physical validation remained essential for accurate results. - The study highlights trade-offs between accessibility of conventional CT imaging and the higher spatial resolution of micro-CT or synchrotron methods, which are often impractical because of jar size, cost, and availability. - Overall, this is reported as the largest systematic comparison to date between tomographic imaging and physical analysis of urned cremations, demonstrating both the potential value of **CT** in pre-excavation assessment and the current need for physical analysis to refine and validate virtual findings.
## Clinical Analysis & Structured Key Points
Virtuality VS Reality: Testing a virtual approach for the exploratory analysis of Bronze Age urns from Northern Italy | PLOS One Browse Subject Areas ? Click through the PLOS taxonomy to find articles in your field. For more information about PLOS Subject Areas, click here . Article Authors Metrics Comments Media Coverage Reader Comments Figures Figures Abstract Cremation is among the most destructive funerary practices. Over the past 30 years, radiographic imaging has increasingly been used to non-destructively extract information from cremation burials. Advances in scanners and imaging software have enhanced the quality of data retrievable through this method. However, scanning cremated remains, particularly when urned, presents challenges due to their fragility, size, material density overlaps, and the high cost of imaging equipment. These factors often result in low-quality images where distinguishing different materials becomes difficult and time-consuming. This exploratory research aims to provide and assess qualitative and quantitative information from tomographic images on burials’ taphonomy, biological profiles (sex and age at death), Minimum Number of Individuals (MNI) and thermal alterations to the bones and to evaluate a relatively simple segmentation strategy to analyse the spatial distribution of the remains in the vessels. For this purpose, 10 non-micro-excavated urns from the Middle-Late Bronze Age necropolis of Vicofertile (Parma, Italy) were scanned using Computerised Tomography (CT), subsequently micro-excavated, and the remains analysed to compare the information obtained from the two approaches. When extracted, data about the age classes, burial taphonomy, and heat-induced changes to the bones (e.g., fractures) were confirmed by the physical analysis, while the MNI, hypothesised to be 10 from the scans, was increased by one. Sex estimation was not possible from the scans. The vertical and horizontal distribution of the remains was estimated for cranial, long, spongy bones, upper, lower limbs, and trunk elements, with the vertical cranial distribution being the most accurately predicted category. Overall, the physical analysis proved essential to refine the results of virtual analysis and deal with the drawbacks of low-quality images. Despite the limitations discussed here, this research presents the largest systematic comparison of tomographic imaging and physical analysis of urned cremations to date, highlighting both the potential and the current limitations of CT-based investigation. Citation: Scalise LM, Morigi MP, Cavazzuti C, Brancaccio R, Crema ER, Seracini M, et al. (2026) Virtuality VS Reality: Testing a virtual approach for the exploratory analysis of Bronze Age urns from Northern Italy. PLoS One 21(9): e0358226. https://doi.org/10.1371/journal.pone.0358226 Editor: Carlos P. Odriozola, Universidad de Sevilla, SPAIN Received: April 22, 2026; Accepted: August 28, 2026; Published: September 22, 2026 Copyright: © 2026 Scalise et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: The data and scripts that support the findings of this study are openly available in Zenodo (DOI: 10.5281/zenodo.15042290 ). Full link: https://zenodo.org/records/15042290 . Funding: This work is part of the corresponding author’s (LMS) PhD project funded by the Cambridge Trust International Scholarship ( https://www.cambridgetrust.org/ ), with the contribution of the Robert Sloley Travel Grant (St John’s College, Cambridge), Worts Travelling Scholars Fund and University Fieldwork Fund (Department of Archaeology, University of Cambridge). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have no competing interests to declare. 1. Introduction Fragmentation is one of the intrinsic characteristics of burned bones (including cremations). The excavation and subsequent handling of these remains by archaeologists and anthropologists during sieving, washing, and analysis usually further exacerbate this fragmentation [ 1 – 3 ]. Previous studies have used radiographic images to gain as much data as possible from cremations before inevitably causing further fragmentation during the remains’ extraction [ 4 – 16 ]. Since its first use, scholars have acknowledged the usefulness of this inspection in identifying the presence of objects inside the urns, locating the remains, and facilitating their excavation [ 4 ]. The technological improvement of the scanners and imaging software has further enriched the quantity and quality of the information retrieved [ 10 ]. Scholars have acquired data about the vessels’ preservation, presence of metallic objects, bone fragmentation, bioturbation [ 6 , 7 , 12 , 13 , 17 – 19 ], and identification of anatomical structures in the assemblage [ 8 , 16 , 20 ]. Moreover, the volumetric shrinkage of the remains [ 21 ], the deposition pattern of the bones, and measurement of their total volume before and after the excavation have been calculated [ 6 , 9 , 14 ]. Some studies have carried out a more holistic analysis of cremation urns combining virtual and physical approaches. They obtained data regarding the taphonomy of the combustion and burial, grave goods, fragmentation rate of bones, and estimation of the sex and age at death of the deceased [ 5 , 10 , 16 , 20 ]. Despite this progress, the virtual analysis of cremated remains has some limitations. The similar densities of the materials and resulting grey values overlap, making their distinction in the images very difficult. Consequently, the segmentation process (i.e., the virtual characterisation of materials based on density values) becomes time-consuming [ 4 , 5 , 8 – 10 , 13 , 16 ]. Unfortunately, this step is the basis of quantitative analysis (e.g., volumetric and morphometric analyses). To overcome this problem, some researchers have tried MRI (Magnetic Resonance Imaging), which enhances the contrast between bone and other materials embedding it. However, Computerised Tomography (CT) scans seemed more spatially accurate [ 22 ]. Gastelum-Strozzi and colleagues [ 6 ] tried to develop a semi-automated framework for processing CT data to make it replicable and less time-consuming. Unfortunately, they used a complex approach that would require considerable training. Waltenberger and colleagues [ 16 ] used a semi-automatic segmentation averaging the Hounsfield units extracted from different materials. However, a manual correction was often necessary due to the presence of artifacts in the scans, and a few bones could be recognised. Furthermore, the jars’ average size generally prevents using more powerful machines, like µ-CT scanners, that would provide higher resolution images and would simplify the selection of the bones. Alternatives, such as facilities for analysing construction materials [ 10 ] or synchrotron facilities [ 23 ], exist but are costly and often unavailable locally. This article presents a novel, simple and relatively quick approach to conducting quantitative and qualitative analysis of urned cremated remains using average-quality tomographic images. The main purpose of this exploratory analysis is extracting information about 1) the urns’ taphonomy, including the impact of factors such as post-depositional disturbance, soil compression, and possible diagenetic alterations 2) the individuals’ sex and age at death, 3) the Minimum Number of Individuals, 4) the thermal alterations to the bones, and 5) the distribution of remains inside the vessels. This last goal would be relevant to identifying potential patterns which might suggest intentional choices and provide insights into funerary behaviours. All these objectives are common to both funerary archaeology and osteological research, regardless of the archaeological context, as they contribute to reconstructing the biological profile of past populations and investigating funerary practices. The present study does not aim to address these archaeological questions in detail. Rather, its objective is methodological: to evaluate a relatively simple approach that maximises the amount of information that can be reliably extracted from average-quality tomographic images to investigate these established research questions. By integrating and validating the virtual procedure through traditional physical analysis, the proposed workflow enabled the retrieval of valuable data on the biological profile and funerary practices of a Late Bronze Age Terramare community (Northern Italy, 1650−1150 BCE). Although validated using this archaeological context, the methodology is applicable to funerary contexts from any period or region. The broader significance of these findings within the archaeological context and their implications for interpreting the community’s life and death will be explored separately. This research primarily focuses on presenting and discussing the analytical approach. 2. Materials and methods 2.1. Archaeological background and sample selection A group of 10 urns ( S1 Table ) was selected from 35 burials excavated in 2009 at Vicofertile, a necropolis discovered during residential development southwest of Parma (Emilia-Romagna, Italy) ( Fig 1 ). All necessary permits were obtained for the described study by the Superintendence of Archaeology, Fine Arts, and Landscape for the provinces of Parma and Piacenza, which complied with all relevant regulations. The remains are currently stored and curated at the Pilotta Monumental Complex in Parma. This research received a favourable ethical opinion from the Department of Archaeology Ethics Board (reference number: ARCH-01-2022-07) at the University of Cambridge (Cambridge, UK). The relative chronology, based on urn and lids shapes and decorations, as well as grave good typology, suggests dating to the advanced Middle Bronze Age (1450/1400–1330/1300 BCE) or the Recent Bronze Age onset (1330/1300–1225/1200 BCE) [ 24 ]. As in only a few other Terramare necropolises in the southern Po Plain (e.g., Casinalbo), burial nuclei of varying size were identified at Vicofertile as part of the internal organisation of the cemetery [ 25 ]. Two vessels were remarkably buried at the centre of separate tumuli, representing the first evidence of tumulus use in this culture’s necropolises [ 24 ] ( Fig 1 ). The Terramare culture spread between 1650 BCE and 1150 BCE across the central Po Plain [ 26 – 30 ], with around 200 settlements and a dozen cemeteries identified across Lombardy, Veneto and Emilia [ 27 , 29 , 30 ]. Download: PNG larger image TIFF original image Fig 1. The location of Vicofertile in the proximity of the Po River (Parma, Italy) (left). Map of Vicofertile (right). The map of Italy is made with Natural Earth (Free vector and raster map data @ naturalearthdata.com). The numbered black spots on the map of Vicofetile represent the burials. The circled spots show the burials selected for CT scanning. The map of the site is adapted from Ferrari and Mutti [ 24 ] and is provided for illustrative purposes only. https://doi.org/10.1371/journal.pone.0358226.g001 The selection of urns for the present study was based on the dimensions and location of the vessels. They showed a good state of preservation to be moved, as they were bounded and excavated en-bloc . The vessels buried at the centre of the tumuli (i.e., burial 15 and 29) and the smallest urn (i.e., burial 26 with a maximum circumference of 38 cm) were included. Otherwise, samples from multiple areas of the site were chosen to equally represent the assemblage ( Fig 1 ). 2.2. Computerised tomography scans The number of scans (i.e., 10) was established according to the time available at the X-ray Computed Tomography Laboratory in Ravenna (University of Bologna), where the urns were scanned. The flexible scanning system, designed for cultural heritage objects, includes a cone-beam X-ray source (Smart EVO 200D, 30–200 kV, 6 mA max), an X-ray detector (Hamamatsu C10900D, CsI: TI scintillator, 1216 × 1232 pixel, 100 μm pixel size), and a rotational sample plate. The vessels were scanned with a voltage of 200 kV (the maximum reachable by the X-ray tube), a current of 3.5 mA, and a 1 mm iron plate collimated with a lead mask to reduce X-ray scattering. All scans were completed in two steps (450 projections, 180° angle per step) around 50 minutes each, as the X-ray source could not run continuously, at one-frame average and exposure times of 4.6–5.2 seconds, achieving a final voxel size of 253 μm. Burial 15, with the largest diameter (27 cm) and a prominent position in the largest tumulus, required a modified protocol. The field of view was reduced to concentrate ray power, and the urn was scanned in four steps (two for the upper and two for the lower half). Sections were merged during post-processing using overlapping reference points. This effective but time-intensive method took approximately five hours, making it unsuitable for all scans. 2.3. Images post-processing, inspection and segmentation strategy The reconstruction and correction of tomographic images were carried out using the software PARREC [ 31 ], implementing the Feldkamp, Davis, and Kress (FDK) cone-beam algorithm [ 32 ]. The post-processing of the images removed artefacts (e.g., outliers, rings caused by X-ray interactions) and the Beam Hardening Correction (BHC) was applied to address selective photon attenuation [ 33 ]. The image stacks were adjusted (i.e., alteration of histogram values, brightness, contrast) and filtered with ImageJ’s unsharp mask [ 34 ]. The adjusted stacks were imported in Avizo 2022.2 (Thermo Fisher Scientific) for visual examination along XY, XZ, YZ axes. This included assessing the presence of lids, grave goods, urn preservation, compression of remains, and bone fractures (e.g., mosaic, longitudinal, bulls-eye) [ 35 – 37 ]. Observations relevant to estimating sex, age-at-death (methods in 2.4), and anatomical elements were recorded. A segmentation strategy was developed to register and quantify the distribution of the remains inside the vessel. Due to the poor contrast between soil, bone, and ceramics, automatic segmentation was unfeasible. The manual segmentation of each slice would have been extremely time-consuming, so six slices for each view were selected and then semi-automatically or manually segmented. Therefore, 18 slices per urn were processed. This sampling was based on the identification of the general range of slices containing recognisable remains from the top to the bottom of the urn. Then, five equal intervals of slices were calculated, and the six slices delimiting the intervals were segmented ( Fig 2 ). This procedure was repeated for each view (XY-XZ-YZ). Depending on the quantity of visible remains, the segmentation took 3–5 days per urn. The number of segmented slices was selected according to the urn’s diameter and the height of the bone deposit, with the intent of keeping the interval between slices at roughly 3.5–4 cm. This spacing was chosen to minimise the risk of overlooking diagnostically relevant elements, which typically measure more than 2–3 cm. Smaller components, such as teeth or phalanges, may not always be captured within a single slice, but they should remain identifiable through qualitative assessment of the CT images. The total number of slices can be modified if a more detailed spatial distribution of elements is needed or in cases where larger urn dimensions require increased sampling resolution. Download: PNG larger image TIFF original image Fig 2. Schematic drawing of slices selection for segmentation and definition of the areas of the vessel. Upper, central, and lower are the vertical sections, while Half A, B, C, and D represent the horizontal sections. https://doi.org/10.1371/journal.pone.0358226.g002 Specific anatomical elements were individually labelled when they could be clearly recognised (e.g., tooth, femoral epiphysis). All other segmented bones were grouped into broader categories termed ‘cranial’, ‘spongy’, and ‘long bone’. For the purposes of statistical analysis, three additional classes, ‘trunk’, ‘upper limbs’, and ‘lower limbs’, were introduced, and identified elements were assigned to the most appropriate of these groups (e.g., a femoral epiphysis was placed within the ‘lower limbs’ class) ( Fig 3 ). Elements segmented as generic ‘epiphyses’ that could not be confidently attributed to either the upper or lower limbs were included within the more general ‘spongy bone’ category. Clavicles and scapulae were placed in the ‘upper limbs’ group, and coxal bones within the ‘lower limbs’ ( Table 1 ). Although this grouping may appear anatomically imprecise, it was considered inappropriate to create separate small categories for structures such as the scapular girdle, pelvic girdle, or unassigned epiphyses. Introducing additional groups would have increased the number of variables and resulted in an uneven dataset, with some categories containing only a few elements and others considerably more, which would have placed unnecessary strain on the statistical analysis. Therefore, bones were assigned either to the most similar category in terms of tissue characteristics (e.g., spongy bone and epiphyses) or to the group representing the closest anatomical association (e.g., coxal bones with the lower limbs as components of the appendicular skeleton). Download: PNG larger image TIFF original image Table 1. Classification of anatomical categories and corresponding skeletal elements used in this study. https://doi.org/10.1371/journal.pone.0358226.t001 Download: PNG larger image TIFF original image Fig 3. Example of different anatomical categories segmented (burial 26). https://doi.org/10.1371/journal.pone.0358226.g003 Following segmentation, statistical measurements describing the spatial distribution of the remains were generated in Avizo 2022 2.2. The area (3D) and volume (3D) occupied by each anatomical class (cranial, spongy, long bones, trunk, upper limbs, and lower limbs) were calculated within the upper, middle, and lower sections of each vessel. Although segmentation is performed on individual slices, the resulting measurements are always volumetric, as each slice has a thickness defined by the voxel size (0.253 mm), and the mesh is produced by triangulating the vertices of these volumetric voxels. To define the vertical sections, the total number of slices containing identifiable remains was divided into three equal parts. For the horizontal analysis, the central XZ and YZ planes were used to divide the distribution into halves ( Fig 2 ). Volume and area values were then recorded to quantify the horizontal distribution of each anatomical category across the vessel. 2.4. Physical excavation and analysis The micro-excavation of the urns was conducted at the Bones Lab and ArcheoLaBio, University of Bologna. The scans were employed as guides during the process. Sagittal images were used to estimate the thickness of soil and bone layers. The measure of the latter was divided by three to calculate the upper, central and lower areas for comparability between the virtual and physical layers. The excavation progressed in these three artificial layers. Each layer was also subdivided in four quarters, to register the horizontal distribution of the remains, comparable to the virtual halves. At the beginning of the micro-excavation, the vessel was rotated so that its position corresponded to the orientation of the virtual model. This alignment ensured that the physical quadrants defined during excavation matched the virtual halves identified in the digital reconstruction. Up to 1 cm variance between virtual and physical layers was considered acceptable, as m
## Related Clinical Research

- [Rising prevalence of single and multiple organ fibrosis in England, 2012–2022](https://medichelpline.com/clinical-feed/bmj-open-2-temporal-trends-in-single-and-multiple-organ-fibrosis-prevalence-and-primary.md)
- [First‑trimester glycolipid and inflammatory markers linked to gestational diabetes risk in rural S](https://medichelpline.com/clinical-feed/bmj-open-11-evaluation-of-the-first-trimester-maternal-glycolipid-profile-on-gestational.md)
- [Problem-based learning improves training outcomes for Central Sterile Supply Department nurses](https://medichelpline.com/clinical-feed/plos-one-5-effect-of-problem-based-learning-in-the-training-of-central-sterile-supply.md)
- [Maternal Vitamin D Status and Preeclampsia Risk in Kenyan Pregnant Women: A Matched Case–Control S](https://medichelpline.com/clinical-feed/plos-one-12-maternal-serum-vitamin-d-levels-and-preeclampsia-among-kenyan-pregnant-women-a.md)
- [Generative Neuromorphic Programming of Mammalian Cells Using ERNs and Biomorphic Neural Networks](https://medichelpline.com/clinical-feed/biorxiv-0-generative-neuromorphic-programming-of-mammalian-cells.md)

## Navigation
- [← Back to General Feed](https://medichelpline.com/clinical-feed/general.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.