Paleoanthropology Group, Department of Paleobiology, Museo Nacional de Ciencias Naturales (MNCN-CSIC), Madrid, Spain.
Department of Biology, Faculty of Sciences, Universidad Autónoma de Madrid (UAM), Madrid, Spain.
J Anat. 2024 Jul;245(1):27-34. doi: 10.1111/joa.14022. Epub 2024 Feb 5.
Up to now, there have been no publication standardizing the digital reconstruction of the modern human ribcage from commingled costo-vertebral material. Consequently, we designed a validated protocol based on anatomical features observed in the literature and the CT scanned ribcages of 10 adult European individuals. After quantifying the shape of these ribcages using 3D geometric morphometrics, we split each vertebra and rib within their corresponding (semi)landmarks. Subsequently, individual bones + (semi)landmarks were imported to LhpFusionBox, commingled and 3D reconstructed. To validate the accuracy of the protocol, we first reconstructed a randomly chosen ribcage three times and then compared these reconstructions to the rest of the sample. Since these reconstructions were closer to their original counterpart than to the others, the remaining sample was reconstructed once. Next, we tested the intra-observer error during reconstructing using the Procrustes distances among the original ribcages and the reconstructions. We observed that first each ribcage reconstruction was clustered to its original counterpart and second there was a learning curve showing an improvement in the reconstruction process over time. Subsequently, we explored general size and shape differences among the original and reconstructed ribcages through a study of centroid size and a permutation test on the Procrustes distances (10,000 permutations), respectively. Specific shape differences between both groups were further examined through a principal component analysis in shape space. None of these analyses found statistical differences between the original and reconstructed ribcages (p > 0.05). Eventually, we extracted the mean shapes of the original ribcages and the reconstructions in order to visualize potential deviations caused by the anatomical considerations of the researcher. These results demonstrate that the protocol is accurate enough to be used when reconstructing a disarticulated human ribcage.
迄今为止,尚无出版物规范从混合的肋-椎材料中对现代人类胸廓进行数字重建。因此,我们根据文献中观察到的解剖学特征和 10 个欧洲成年人的 CT 扫描肋骨设计了一个经过验证的方案。在使用 3D 几何形态测量量化这些胸廓的形状后,我们将每个椎骨和肋骨分割到其相应的(半)地标内。随后,将单个骨骼+(半)地标导入 LhpFusionBox 中,混合并进行 3D 重建。为了验证方案的准确性,我们首先随机重建了一个肋骨,然后将这三个重建结果与其余样本进行比较。由于这些重建结果与其原始样本更接近,而与其他样本相差较大,因此仅对其余样本进行了一次重建。接下来,我们通过原始肋骨与重建肋骨之间的 Procrustes 距离来测试在重建过程中的观察者内误差。我们观察到,首先,每个胸廓重建都聚类到其原始样本,其次,随着时间的推移,重建过程中存在学习曲线,表明重建效果不断提高。随后,我们通过研究质心大小和对 Procrustes 距离进行置换检验(10,000 次置换),分别从原始和重建肋骨的整体大小和形状差异进行了探索。通过形状空间中的主成分分析进一步检查了两组之间的特定形状差异。这些分析均未发现原始和重建肋骨之间存在统计学差异(p>0.05)。最后,我们提取了原始和重建肋骨的平均形状,以可视化由于研究人员的解剖学考虑而导致的潜在偏差。这些结果表明,该方案足够准确,可以在对分离的人类胸廓进行重建时使用。