Exploring palatal and dental shape variation with 3D shape analysis and geometric deep learning

Abstract

Objectives: Palatal shape contains a lot of information that is of clinical interest. Moreover, palatal shape analysis can be used to guide or evaluate orthodontic treatments. A statistical shape model (SSM) is a tool that, by means of dimensionality reduction, aims at compactly modeling the variance of complex shapes for efficient analysis. In this report, we evaluate several competing approaches to constructing SSMs for the human palate. Setting and Sample Population:This study used a sample comprising digitized 3D maxillary dental casts from 1,324 individuals. Materials and methods: Principal component analysis (PCA) and autoencoders (AE)are popular approaches to construct SSMs. PCA is a dimension reduction technique that provides a compact description of shapes by uncorrelated variables. AEs are situated in the field of deep learning and provide a non-linear framework for dimension reduction. This work introduces the singular autoencoder (SAE), a hybrid approach that combines the most important properties of PCA and AEs. We assess the performance of the SAE using standard evaluation tools for SSMs, including accuracy, generalization, and specificity.

Wear patterns in knee OA correlate with native limb geometry

“…Linearly transforming the data in a new coordinate system, variation in the data is captured in multiple principal components. The previously obtained joint space images, however, are inherently non-linear and are as such not amenable for analysis by classical linear techniques including PCA (Nauwelaers et al, 2021). Nevertheless, this type of data is ideal for use in neural-based learning.…”

“…Progress in computer vision and image analysis now offers efficient methods to establish anatomical correspondence between knee shapes and to describe the joint space to perform statistical comparison and pattern analysis of sets of joint space narrowing maps (i.e., convolutional neural networks following conformal mesh parameterization) (Audenaert et al, 2019a;Nauwelaers et al, 2021). Whereas Favre and colleagues aimed for anatomical correspondence relying on standardized two-dimensional pixel-maps, anatomical correspondence in 3D can be obtained by nonrigid surface registration of a reference template thereby providing a dense set of homologous landmarks amenable to statistical analysis (Williams et al, 2010;Favre et al, 2013;Audenaert et al, 2019a;van Houcke et al, 2020;Peiffer et al, 2022).…”

“…Whereas Favre and colleagues aimed for anatomical correspondence relying on standardized two-dimensional pixel-maps, anatomical correspondence in 3D can be obtained by nonrigid surface registration of a reference template thereby providing a dense set of homologous landmarks amenable to statistical analysis (Williams et al, 2010;Favre et al, 2013;Audenaert et al, 2019a;van Houcke et al, 2020;Peiffer et al, 2022). This permits statistical comparison between and within populations of skeletal shape and related joint space geometry (Williams et al, 2010;Audenaert et al, 2019a;Nauwelaers et al, 2021). Furthermore and by comparison with healthy data, disease severity and progression can be quantified and even a virtual healthy twin can be generated (Fuessinger et al, 2019;Ahmadian et al, 2022).…”

Artificial intelligence and machine learning in orthodontics

“…Although essential for the success of AI-based diagnosis, the objective extraction of these morphological features is not easy. Nauwelaers et al developed a novel AIbased method for the analysis of palatal 3D shape 26 and Croquet et al developed methods of automatic landmarking for analysis of the palatal shape. 27 Machine/deep learning is a powerful tool for assessment of important clinical parameters and the extraction of clinical knowledge from complex patient information and data that affects the treatment outcomes.…”

“…Although essential for the success of AI‐based diagnosis, the objective extraction of these morphological features is not easy. Nauwelaers et al developed a novel AI‐based method for the analysis of palatal 3D shape 26 and Croquet et al developed methods of automatic landmarking for analysis of the palatal shape 27 …”