Domen Močnik1, Bulat Ibragimov2, Lei Xing3, Primož Strojan4, Boštjan Likar5, Franjo Pernuš6, Tomaž Vrtovec7. 1. Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia. Electronic address: domen.mocnik@fe.uni-lj.si. 2. Department of Radiation Oncology, Stanford University School of Medicine, Stanford, CA, USA. Electronic address: bulat@stanford.edu. 3. Department of Radiation Oncology, Stanford University School of Medicine, Stanford, CA, USA. Electronic address: lei@stanford.edu. 4. Institute of Oncology Ljubljana, Ljubljana, Slovenia. Electronic address: primoz.strojan@onko-i.si. 5. Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia. Electronic address: bostjan.likar@fe.uni-lj.si. 6. Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia. Electronic address: franjo.pernus@fe.uni-lj.si. 7. Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia. Electronic address: tomaz.vrtovec@fe.uni-lj.si.
Abstract
PURPOSE: To develop an automatic multimodal method for segmentation of parotid glands (PGs) from pre-registered computed tomography (CT) and magnetic resonance (MR) images and compare its results to the results of an existing state-of-the-art algorithm that segments PGs from CT images only. METHODS: Magnetic resonance images of head and neck were registered to the accompanying CT images using two different state-of-the-art registration procedures. The reference domains of registered image pairs were divided on the complementary PG regions and backgrounds according to the manual delineation of PGs on CT images, provided by a physician. Patches of intensity values from both image modalities, centered around randomly sampled voxels from the reference domain, served as positive or negative samples in the training of the convolutional neural network (CNN) classifier. The trained CNN accepted a previously unseen (registered) image pair and classified its voxels according to the resemblance of its patches to the patches used for training. The final segmentation was refined using a graph-cut algorithm, followed by the dilate-erode operations. RESULTS: Using the same image dataset, segmentation of PGs was performed using the proposed multimodal algorithm and an existing monomodal algorithm, which segments PGs from CT images only. The mean value of the achieved Dice overlapping coefficient for the proposed algorithm was 78.8%, while the corresponding mean value for the monomodal algorithm was 76.5%. CONCLUSIONS: Automatic PG segmentation on the planning CT image can be augmented with the MR image modality, leading to an improved RT planning of head and neck cancer.
PURPOSE: To develop an automatic multimodal method for segmentation of parotid glands (PGs) from pre-registered computed tomography (CT) and magnetic resonance (MR) images and compare its results to the results of an existing state-of-the-art algorithm that segments PGs from CT images only. METHODS: Magnetic resonance images of head and neck were registered to the accompanying CT images using two different state-of-the-art registration procedures. The reference domains of registered image pairs were divided on the complementary PG regions and backgrounds according to the manual delineation of PGs on CT images, provided by a physician. Patches of intensity values from both image modalities, centered around randomly sampled voxels from the reference domain, served as positive or negative samples in the training of the convolutional neural network (CNN) classifier. The trained CNN accepted a previously unseen (registered) image pair and classified its voxels according to the resemblance of its patches to the patches used for training. The final segmentation was refined using a graph-cut algorithm, followed by the dilate-erode operations. RESULTS: Using the same image dataset, segmentation of PGs was performed using the proposed multimodal algorithm and an existing monomodal algorithm, which segments PGs from CT images only. The mean value of the achieved Dice overlapping coefficient for the proposed algorithm was 78.8%, while the corresponding mean value for the monomodal algorithm was 76.5%. CONCLUSIONS: Automatic PG segmentation on the planning CT image can be augmented with the MR image modality, leading to an improved RT planning of head and neck cancer.
Authors: David Mattes; David R Haynor; Hubert Vesselle; Thomas K Lewellen; William Eubank Journal: IEEE Trans Med Imaging Date: 2003-01 Impact factor: 10.048
Authors: E Faggiano; C Fiorino; E Scalco; S Broggi; M Cattaneo; E Maggiulli; I Dell'Oca; N Di Muzio; R Calandrino; G Rizzo Journal: Phys Med Biol Date: 2011-01-14 Impact factor: 3.609
Authors: Charlotte L Brouwer; Roel J H M Steenbakkers; Edwin van den Heuvel; Joop C Duppen; Arash Navran; Henk P Bijl; Olga Chouvalova; Fred R Burlage; Harm Meertens; Johannes A Langendijk; Aart A van 't Veld Journal: Radiat Oncol Date: 2012-03-13 Impact factor: 3.481