Reliability and Agreement of Automated Head Measurements From 3-Dimensional Photogrammetry in Young Children

Abstract

This study aimed to assess the reliability and agreement of automated head measurements using 3-dimensional (3D) photogrammetry in young children. Specifically, the study evaluated the agreement between manual and automated occipitofrontal circumference (OFC) measurements (n = 264) obtained from 3D images of 188 patients diagnosed with sagittal synostosis using a novel automated method proposed in this study. In addition, the study aimed to determine the interrater and intrarater reliability of the automatically extracted OFC, cephalic index, and volume. The results of the study showed that the automated OFC measurements had an excellent agreement with manual measurements, with a very strong regression score (R 2 = 0.969) and a small mean difference of −0.1 cm (−0.2%). The limits of agreement ranged from −0.93 to 0.74 cm, falling within the reported limits of agreement for manual OFC measurements. High interrater and intrarater reliability of OFC, cephalic index, and volume measurements were also demonstrated. The proposed method for automated OFC measurements was found to be a reliable alternative to manual measurements, which may be particularly beneficial in young children who undergo 3D imaging in craniofacial centers as part of their treatment protocol and in research settings that require a reproducible and transparent pipeline for anthropometric measurements. The method has been incorporated into CraniumPy, an open-source tool for 3D image visualization, registration, and optimization, which is publicly available on GitHub (https://github.com/T-AbdelAlim/CraniumPy).

3D Analysis of the Cranial and Facial Shape in Craniosynostosis Patients: A Systematic Review

“…Alternatively, the level corresponding to maximum head length has been suggested for automated extraction of cephalic measurements. This method has shown to be more robust as it circumvents slices containing the ears, thus simplifying the measurement process 12 …”

“…Of these 13 publications, 10 used a measurement plane parallel to the base plane 11,12,14,22,24,25,30–32,45 . The base plane was mostly created using the landmarks left and right tragus and nasion/sellion 11,12,14,22,24,45 . In addition, other publications used the Frankfort horizontal plane, 32 a plane through the left and right tragus, subnasale, 25 left and right exocanthions, 30 or the nasion and clinoid processes dorsum sella 31 .…”

“…In addition, other publications used the Frankfort horizontal plane, 32 a plane through the left and right tragus, subnasale, 25 left and right exocanthions, 30 or the nasion and clinoid processes dorsum sella 31 . The measurement plane was then created parallel to the base plane at a certain level, with the majority of the publications at the level of maximum circumference 11,12,30–32 …”

“…Of those 24 publications, 19 publications were considered as reproducible and are described in Supplemental Digital Content Table 2 (http://links.lww.com/SCS/F940). Fourteen publications defined ICV as the volume above a certain base plane 11–28 . In 1 publication, the cranium above the base plane was divided into 10 proportionally spaced intervals, with level 0 at base plane and level 10 at the vertex 14 .…”

“…All publications used landmarks to create the base plane. Seven publications used both tragus and the nasion/sellion, 11,12,14,20,22,24,27 and 4 publications used both tragus and the subnasale point 19,25,26,29 . Three publications used the tragus and the lateral canthi as landmarks 21,23,28 .…”