Our research group, is developing a completely automatic pipeline that is able to create patient specific FE models from radiological data sets, using patient specific geometries and materials. The FE simulations will help the surgeons at Balgrist hospital to decide what the ideal operation for the individual patient is.
However, at the moment the link is missing to degenerative diseases that may affect the spine severely. During the last years, machine learning tools such as tensor flow have become very powerful in automatic feature detection and autonomous segmentation.The idea of the project is hence, to generate a platform that is able to detect spinal pathologies, such as disc herniation’s, disc degeneration, modic changes, ect. from MRI. Furthermore, automatic segmentation and extraction of the 3D-disc geometry is also a requirement for the software. The platform will be developed in Matlab and Python and uses Tensor Flow for deep learning. The project is performed in close interaction with the surgeons who assess the spines from the MRI and deliver the data sets for training.
Tasks:
- A Matlab and Phyton based environment is developed that is able to detect spinal pathologies and degenerative changes.
- 3D segmentation of disc geometry on MRI using a tensor flow based U-net.
- The project takes place in close interaction with the surgeons who assess the spines on the radiological data sets
Requirements:
• Solid Matlab skills
• Intrinsic motivation to work on a project that supports surgeons in their work of treating patients
Our research group, is developing a completely automatic pipeline that is able to create patient specific FE models from radiological data sets, using patient specific geometries and materials. The FE simulations will help the surgeons at Balgrist hospital to decide what the ideal operation for the individual patient is. However, at the moment the link is missing to degenerative diseases that may affect the spine severely. During the last years, machine learning tools such as tensor flow have become very powerful in automatic feature detection and autonomous segmentation.The idea of the project is hence, to generate a platform that is able to detect spinal pathologies, such as disc herniation’s, disc degeneration, modic changes, ect. from MRI. Furthermore, automatic segmentation and extraction of the 3D-disc geometry is also a requirement for the software. The platform will be developed in Matlab and Python and uses Tensor Flow for deep learning. The project is performed in close interaction with the surgeons who assess the spines from the MRI and deliver the data sets for training.
Tasks:
- A Matlab and Phyton based environment is developed that is able to detect spinal pathologies and degenerative changes.
- 3D segmentation of disc geometry on MRI using a tensor flow based U-net.
- The project takes place in close interaction with the surgeons who assess the spines on the radiological data sets
Requirements:
• Solid Matlab skills
• Intrinsic motivation to work on a project that supports surgeons in their work of treating patients
Development of a platform for automatic detection of spinal pathologies and segmentation of disc in 3D.
Development of a platform for automatic detection of spinal pathologies and segmentation of disc in 3D.
jonas.widmer@hest.ethz.ch
(if you are applying for an internship, please note that we cannot provide a salary since we are a university department)
jonas.widmer@hest.ethz.ch (if you are applying for an internship, please note that we cannot provide a salary since we are a university department)