Combining deep learning with 3D stereophotogrammetry for craniosynostosis diagnosis

Abstract

Craniosynostosis is a condition in which cranial sutures fuse prematurely, causing problems in normal brain and skull growth in infants. To limit the extent of cosmetic and functional problems, swift diagnosis is needed. The goal of this study is to investigate if a deep learning algorithm is capable of correctly classifying the head shape of infants as either healthy controls, or as one of the following three craniosynostosis subtypes; scaphocephaly, trigonocephaly or anterior plagiocephaly. In order to acquire cranial shape data, 3D stereophotographs were made during routine pre-operative appointments of scaphocephaly (n = 76), trigonocephaly (n = 40) and anterior plagiocephaly (n = 27) patients. 3D Stereophotographs of healthy infants (n = 53) were made between the age of 3–6 months. The cranial shape data was sampled and a deep learning network was used to classify the cranial shape data as either: healthy control, scaphocephaly patient, trigonocephaly patient or anterior plagiocephaly patient. For the training and testing of the deep learning network, a stratified tenfold cross validation was used. During testing 195 out of 196 3D stereophotographs (99.5%) were correctly classified. This study shows that trained deep learning algorithms, based on 3D stereophotographs, can discriminate between craniosynostosis subtypes and healthy controls with high accuracy.

Spherical harmonics to quantify cranial asymmetry in deformational plagiocephaly

“…In spite of the fact that CT and MRI are considered the gold standard for diagnostics and surgery intervention, derived measurements are usually obtained manually by radiologists. Because of this, it is required to improve and develop new, mathematically validated anthropometric indexes, that provide useful and accurate information to specialised doctors 14 , 42 .…”

“…The methodologies to obtain three-dimensional (3D) models include low-cost tools 12 and, more commonly, setups of cameras and scanners 13 , 14 . Radiological tests, such as Computed Tomography and Magnetic Resonance Imaging are considered the gold standard but they are also highly invasive and costly.…”

“…Some authors have applied different mathematical processes to the classification and evaluation of cranial deformation from 3D models. Some examples are Principal Components 15 , the Root Mean Square differences between head quadrants 16 , kernel density estimation 17 , and deep learning for craniosynostosis diagnosis 14 . Furthermore, cranial deformation has been assessed by comparing the measured head shape to a fitted tri-axial ellipsoid which was considered as an ideal shape for a non-deformed cranium 18 .…”

The Characterisation of the Craniofacial Morphology of Infants Born With Zika Virus; Innovative Approach for Public Health Surveillance and Broad Clinical Applications

“…The 3D stereophotogrammetry system has been applied on the capture and analysis of the craniofacial morphology ( 24 – 30 ); further mobile phone-based system ( 31 ) has been developed. In this paper, a 3D photogrammetry system has been applied on the analysis of the craniofacial morphology of infants born with Zika virus.…”

Multi-Scale Part-Based Syndrome Classification of 3D Facial Images

“…Given the increasing accessibility of 3D imaging hardware, including consumer-grade depth sensors in modern smartphones [7], large-scale 3D shape analysis and deep learning for syndrome classification is becoming a practical possibility. Previous work in this domain has used linear dimensionality reduction and classification techniques, or feed-forward neural networks, and has typically focused on discriminating one or a few syndrome groups from controls [8]- [10]. The most comprehensive attempt at 3D multi-syndrome classification deployed linear techniques on 64 syndrome classes, with a sparse configuration of 65 anatomical 3D facial landmarks [3].…”