[Insight-users] Need for collaboration or courses in medical image processing (with ITK)

[Olivier Sebastien Jerome Cros / Region Nordjylland]

Dear Insight Users,

My name is Olivier Cros, working in Denmark in the hospital of Aalborg (north of Jutland). I am currently doing research on a very complex bone around the ear known as the mastoid bone about its structure and its functionality. I am educated as a Master of Science in Biomedical Engineering with the specialization of Medical Informatics. For some years, I've been extensively working as a research assistant on the development of virtual surgery simulator (human brain ventricles puncture simulator and knee arthroscopy simulation). I decided to work in the local hospital as an engineer to help ear surgeons on the understanding of the mastoid bone (as explained above).

The mastoid bone contains numerous and complex ramifications called air cells which are supposed to be connected together. This structure is very efficient for the purpose of gas exchange similar to the function of the bronchi and alveoli of the lungs. It has also similar properties from a histological point of view, such as for instance the flat epithelium with rich vascularization beneath the capillary network found in the lungs. The mastoid bone is sometimes called the miniature lungs.

My main mission is to extract some quantitative information such as volume and area of the air cells contained in the mastoid bone and compute the volume/area ratio in order to investigate this gas-exchange phenomena. From a medical aspect, otitis media (a very common childhood disease) is, among other sequelae, related to decreased hearing as well as the development of negative middle ear pressure.  This is caused by a net absorption of gas predominantly for the mastoid mucosae.

I have in possession some high resolution CT scans of the temporal bone (DICOM format) I can work with. First I decided to extract a skeleton of the air cells ramifications. After reading some literatures, I worked on a binary thresholding of the temporal bone (after inversion of the binary files) and used the 3D thinning available in the MIDAS repository. My next step is to extract information from the skeleton such as the number of branches, their labeling, the distance of the branches, etc.

So far I've been using ITK for most of the steps I've been through. But I can see that my knowledge in medical image processing is not deep enough and my programming skills are not as good as some of yours people :-). Therefore, I've been looking on different institutes around the world for finding a possible collaboration, a possible education and/or possibly a PhD.

I would be very interested in some courses where I could learn more about medical image processing in combination with ITK. I know that Matlab can do the job as well (used it many times) but my programming knowledge is not sufficient enough to convert Matlab coding and ITK. So far the ITk source code I used is coming from the very good examples from ITK in connection with their book (that I have in possession in hard copy).

Sorry for this very long mail, and if you are interested don't hesitate to contact me. I will be pleased to reply to you.

Thank you in advance,

Best regards,

Olivier Cros.