Bogdan Matuszewski, ADSIP, University of Central Lancashire, United Kingdom
Lik-Kwan Shark, ADSIP, University of Central Lancashire, United Kingdom
Christopher Moore, The North Western Medical Physics, the Christie Hospital NHS Foundation Trust; Manchester Cancer Research Centre (MCRC), United Kingdom
Dave Burton, GERI, Liverpool John Moores University, United Kingdom
Aymeric Histace, ENSEA-CNRS, Cergy-Pontoise University, France
Medical image segmentation and registration; Shape modelling; Tracking; CBCT and Image Reconstruction; PET/CT imaging; Patient positioning and position monitoring; Modelling of tumour growth; Image guided radiation therapy; Navigation for biopsy and brachytherapy seeds implantation.
Scope and Topics
Approximately one in three people will develop cancer at some point in their lives. Technical improvements in diagnosis and treatment have significantly contributed to improved survival in recent years: the 5 year rate is now 50% and the 10 year rate has doubled in the last 30 years. Many important advances in combating cancer have been made due to progress in imaging and medical image analysis. New imaging technologies (e.g. PET/CT, Cone Beam CT, tagged MRI, photoacoustic tomography, optical surface sensing) not only improved diagnostic capabilities enabling detection and tumour localisation with much higher precision, but also improved monitoring of the treatment delivery in chemo/radiotherapy (e.g. in radiotherapy through better patient localisation and position monitoring) as well as aided surgical procedures (e.g. implantation of radioactive seeds in brachytherapy). Many of these image processing techniques have matured and are being routinely used in the clinical practice as for example rigid or affine registration but many others are still active and challenging research areas. These include deformable segmentation and registration often with imposed prior shape and/or topology constraints, navigation and object/organ tracking e.g. in fluoroscopy/endoscopy, Cone Beam tomographic reconstructions from small number of projections, or tissue classification. Many of the method developed for medical image processing applications benefited from progress made in general image processing and computer vision, as for example: PDE based image processing including level set methodology, combinatorial optimisation on graphs, Bayesian methodology including variational and Sequential Monte Carlo formalisms, or representing and processing data on manifolds to name some. On the other hand medical imaging applications offer some of the toughest problems in computer vision and image processing due to low resolution, low-contrast and high-level of distortion as well as random and structural noise in the medical images. Therefore problems originated in medical application domain can stimulate development of new techniques in computer vision and image processing.
The objective of this session will be to report on a current progress made in engineering and computational science related to medical imaging, and image analysis for applications relevant to oncology. A non exclusive list of topics of interest for this session is as follows:
- Image registration
- Shape modelling for organ segmentation
- Detection and classification for tissue categorisation
- Tracking and surgical navigation
- CBCT and Image Reconstruction
- Progress in medical imaging including PET/CT, Cone Beam CT, tagged MRI, photoacoustic tomography, optical surface sensing
- Patient positioning and position monitoring
- Modelling of tumour growth
- Image guided radiation therapy
- Navigation for biopsy and implantation of radioactive seeds in brachytherapy
The proposed session came to existence as a result of a European collaboration on the Engineering and Computational Science for Oncology Network (ECSON) project recently funded by the UK’s Engineering and Physical Sciences Research Council.
All accepted papers will be published in a special section of the proceedings book, under an ISBN reference, and on CD-ROM support.