Intracranial Volume Quantification from 3D Photography

Abstract

3D photography offers non-invasive, radiation-free, and anesthetic-free evaluation of craniofacial morphology. However, intracranial volume (ICV) quantification is not possible with current non-invasive imaging systems in order to evaluate brain development in children with cranial pathology. The aim of this study is to develop an automated, radiation-free framework to estimate ICV. Pairs of computed tomography (CT) images and 3D photographs were aligned using registration. We used the real ICV calculated from the CTs and the head volumes from their corresponding 3D photographs to create a regression model. Then, a template 3D photograph was selected as a reference from the data, and a set of landmarks defining the cranial vault were detected automatically on that template. Given the 3D photograph of a new patient, it was registered to the template to estimate the cranial vault area. After obtaining the head volume, the regression model was then used to estimate the ICV. Experiments showed that our volume regression model predicted ICV from head volumes with an average error of 5.81 ± 3.07% and a correlation (R²) of 0.96. We also demonstrated that our automated framework quantified ICV from 3D photography with an average error of 7.02 ± 7.76%, a correlation (R²) of 0.94, and an average estimation error for the position of the cranial base landmarks of 11.39 ± 4.3mm.

Keywords: 3D photography; computed tomography; intracranial volume quantification; registration.

Fig. 1

Schematic of the proposed framework. ICV: intracranial volume. HV: head volume. The methods used for each of the boxes in this figure are detailed in Section 2 of the paper.

Fig. 2

Cranial vault area estimation for head volume measurement. The green surface represents the patient’s head surface, which can be obtained from CT or 3D photograph. The cranial base (in purple) corresponds to two planes defined by the 4 cranial base landmarks at the nasion, opisthion the two clinoid processes of the dorsum sellae.

Fig. 3

Linear regression model predicting intracranial volume based on the automated head volume quantification from 3D photography.

Fig. 4

Bland-Altman plot comparing the estimated ICV from 3D photography (V_{3D_ICV}) the true ICV from CT (V_{CT_ICV}) for the patients in Φ_pairs.

Fig. 5

Representation of the ICV against the age of the patients. Blue diamonds show the true ICV quantified from CT images for the cases from Φ_pairs. Green squares show the ICV estimated from 3D photography for the cases from Φ_pairs, while red squares show the ICV estimated from 3D photography for the cases in Φ_singles. The blue line represents the age regression function estimated for the true ICV obtained from CT, while the magenta line represents the age regression function for the ICV estimated from 3D photography in both Φ_pairs Φ_singles.