Automatic Segmentation of Mandible from Conventional Methods to Deep Learning—A Review

Abstract

Medical imaging techniques, such as (cone beam) computed tomography and magnetic resonance imaging, have proven to be a valuable component for oral and maxillofacial surgery (OMFS). Accurate segmentation of the mandible from head and neck (H&N) scans is an important step in order to build a personalized 3D digital mandible model for 3D printing and treatment planning of OMFS. Segmented mandible structures are used to effectively visualize the mandible volumes and to evaluate particular mandible properties quantitatively. However, mandible segmentation is always challenging for both clinicians and researchers, due to complex structures and higher attenuation materials, such as teeth (filling) or metal implants that easily lead to high noise and strong artifacts during scanning. Moreover, the size and shape of the mandible vary to a large extent between individuals. Therefore, mandible segmentation is a tedious and time-consuming task and requires adequate training to be performed properly. With the advancement of computer vision approaches, researchers have developed several algorithms to automatically segment the mandible during the last two decades. The objective of this review was to present the available fully (semi)automatic segmentation methods of the mandible published in different scientific articles. This review provides a vivid description of the scientific advancements to clinicians and researchers in this field to help develop novel automatic methods for clinical applications.

Videomics of the Upper Aero-Digestive Tract Cancer: Deep Learning Applied to White Light and Narrow Band Imaging for Automatic Segmentation of Endoscopic Images

“…The development of semantic segmentation AI models for medical image analysis is a field of study that is becoming increasingly widespread, especially for radiologic imaging ( 43 – 45 ). Conversely, the exploitation of these algorithms to investigate videoendoscopic images represents a sphere of research less explored in the literature, as demonstrated by the lack of proper terminology to indicate such a field of interest before our first proposal to identify it as “Videomics” ( 17 ).…”

Establishing a Point-of-Care Virtual Planning and 3D Printing Program

“…20,37 A detailed discussion is beyond the scope of this article; however, segmentation can be performed manually, using automated/semi-automated rule-based techniques, machine learned algorithms, or a combination of the three. [38][39][40][41][42][43][44][45][46][47] Performing manual segmentation is very timeconsuming, as the user is required to mark (with a paintbrush or a lasso) each slice of the image stack, which can contain hundreds of slices. 48 To speed up the process, various automated segmentation algorithms exist with the most common being thresholding, edge detection, and region growing.…”

Verification of De-Identification Techniques for Personal Information Using Tree-Based Methods with Shapley Values

“…Recently, machine-learning techniques have been used for data analysis in a variety of fields, including the education [16,17], energy [18,19], environmental [20,21], medical [22][23][24], and security [25][26][27] fields, to assist professionals in saving time and effort because machine learning can effectively discover nonlinear relationships between dependent and independent variables in numerical datasets. To address classification problems in the medical field, Shailaja et al [28] presented various machine-learning techniques, such as the support vector machine, naive Bayes classification, k-nearest neighbors, and others, to predict various diseases, such as heart disease, breast cancer, diabetes, and thyroid disease.…”