Measurement error of 3D cranial landmarks of an ontogenetic sample using Computed Tomography

Computed Tomography (CT) is a powerful tool in craniofacial research that focuses on morphological variation. In this field, an ontogenetic approach has been taken to study the developmental sources of variation and to understand the basis of morphological evolution. This work aimed to determine measurement error (ME) in cranial CT in diverse developmental stages and to characterize how this error relates to different types of landmarks.

We used a sample of fifteen skulls ranging from 0 to 31 years. Two observers placed landmarks in each image three times. Measurement error was assessed before and after Generalized Procrustes Analysis.

The results indicated that ME is larger in neurocranial structures, which are described mainly by type III landmarks and semilandmarks. In addition, adult and infant specimens showed the same level of ME. These results are specially relevant in the context of craniofacial growth research.

CT images have become a frequent evidence to study cranial variation. Evaluation of ME gives insight into the potential source of error in interpreting results. Neural structures present higher ME which is mainly associated to landmark localization. However, this error is irrespective of age. If landmarks are correctly selected, they can be analyzed with the same level of reliability in adults and subadults.