GSoC 2010 Project Ideas
The ideas bellow were proposed based on several brainstorms. We were also guided by community surveys, where members gave ideas and told their needs regarding medical images.
1. Interaction by Speech Recognition
Brief explanation:Allow users to interact with InVesalius in handsfree mode using Sphinx and GStreamer.
Expected results: The user will be able to turn on the speech recognition feature in InVesalius and it will ask "What can I do to help you?". After opening a project / DICOM files, the user will be able to choose to interact with volume or 2D slices and provide commands such as "rotate", "move" and "zoom". This would be extremely helpful in at least two contexts:
- on the impairment of disabled individuals, allowing them to use InVesalius;
- on medical procedures with the guidance of imaging technologies, when surgeons have to use their hands to operate patients and cannot use the mouse.
Besides Sphinx and GStreamer, the student will also have contact with VTK and wxPython.
Prerequisites:
- Knowledge of Python
Reporter: tatiana.alchueyr
Mentor: tatiana.alchueyr
Backup mentors: paulojamorim, totonixsame
2. PACS-Server Integration
Brief explanation: Integrate InVesalius to a DCM4CHEE server in order to download, search and view DICOM files available remotely
Expected results: DCM4CHEE is a PACS (Picture Archiving and Communications System) server. Its task is to archieve medical imaging files on DICOM (Digital Imaging and Communications in Medicine) file-format. By integrating InVesalius to DCM4CHEE, you will allow users to explore medical exams that are not available locally and to download them. Imagine browsing the local network of a hospital and seeing what patient medical image exams are available so radiologists can page them and surgeons can plan surgery. An implementation would supply InVesalius backend (DICOM-client), including client configuration files, server calls and GUI. If the integration with DCM4CHEE works, InVesalius will be able to be integrated to any PACS network involving other PACS servers as well. Regarding the communication, DCMTK could be used. DCMTK is written in C++, so this is also a perfect opportunity to get your feet on learning how to make a Python binding.
http://dl.dropbox.com/u/4053278/invesalius_trac/gsoc/project_pacs.jpg
Prerequisites:
- Knowledge of Python and enthusiasm for trying new things and pushing the future of medical imaging are required.
- Knowledge of wxPython and DICOM protocol are a plus.
Reporter: Based on group brainstorm, written by tatiana.alchueyr
Mentor: paulojamorim
Backup mentors: tatiana.alchueyr, totonixsame
3. Real-time Communication between InVesalius clients
Brief explanation: Integrate InVesalius clients using Telepathy so they can communicate and exchange data
Expected results: Allow radiologists or a physicians who are analyzing complex medical images using InVesalius to get a second opinion on a diagnose, exchanging messages and data with their colleagues. Data includes screenshots, DICOM files, InVesalius projects and possibly movies. This feature would allow radiologists to make reports remotely and could be even used by radiologists students to exchange opinions with a supervisor. Telepathy was choosen due to its ability of enabling real-time communication via pluggable protocol backends.
Prerequisites:
- Knowledge of Python
Reporter: Based on group brainstorm, written by tatiana.alchueyr
Mentor: totonixsame
Backup mentors: tatiana.alchueyr, paulojamorim
Other ideas
- 2D and 3D Cropping Tools for surgery planning
- Mirroring tool for Prosthesis Planning
- Region Growing and advanced segmentation tools
- Panoramic X-Ray Simulation
