Multi Atlas Segmentation with Active Shape Model Refinement for Multi-Organ Segmentation in Head and Neck Cancer Radiotherapy Planning

Abstract

We describe a segmentation method that was used in the Head and Neck Auto Segmentation Challenge held at the MICCAI 2015 conference. The algorithm consists of two building blocks. First, we employ a multi-atlas segmentation to obtain an initial segmentation for the considered organs at risk. Secondly, we use an Active Shape Model (ASM) segmentation to refine the initial segmentation of some of the organs. Leave-one-out experiments with the training data were used to determine suitable parameters for the individual steps of the segmentation. The ASM refinement resulted in improved segmentation for the optic nerves and submandibular glands, while for the brain stem, parotid glands, chiasm, and mandibular bone, the multi-atlas segmentation was preferable. Our submission achieved the second rank in the challenge.

Robust and fully automated segmentation of mandible from CT scans

“…However, this procedure is not trivial due to large morphological variations among different patients, disease burden, and inevitable image artifacts including dental fillings, orthodontic wires, bands, and braces. Towards a solution to this challenging problem, there have been many attempts in the literature; however, most of the studies that have addressed the mandible segmentation utilize statistical shape models [3]. More recently, machine learning based approaches have been proposed [1] to avoid segmentation step and focus on measuring clinically useful metrics.…”

Automatic segmentation of the mandible from computed tomography scans for 3D virtual surgical planning using the convolutional neural network

“…Conventional approaches for mandible segmentation typically use pixel-based or model-based methods. In the past decade, some traditional semi-automatic and automatic methods have been developed to segment the mandible in CT scans (Gollmer and Buzug 2012, Abdi et al 2015, Albrecht et al 2015, Chen and Dawant 2015, Mannion-Haworth et al 2015, Chuang et al 2017, Torosdagli et al 2017. A statistical shape model for mandible segmentation was presented by Gollmer and Buzug (2012).…”

“…The models are then applied to segmentation of ROIs in CT scans. Albrecht et al employed a multi-atlas segmentation to obtain an initial segmentation and then apply an active shape model (ASM) segmentation to refine the initial segmentation of the organ (Albrecht et al 2015). The performances of the conventional methods, however, are often affected by the noise or metal artifacts in the CT images.…”

Deep Learning-Based Automatic Segmentation of Mandible and Maxilla in Multi-Center CT Images

“…In some studies, atlas-based and statistical model-based methods have been combined with each other or with another method, leading to various other approaches for automatic mandible segmentation. Albrecht et al [16] used a multi-atlas to obtain an initial segmentation of the OAR and an active shape model to refine the initial segmentation. Aghdasi et al [17] employed anatomic landmarks and prior knowledge for segmentation.…”