iSpine: A Innovative Spine Image ProcessingToolkit

A versatile spine modeling toolkit would enable the automatic assessment of gross anomalies and the evaluation of limited regions and spine components such as vertebra and discs. The system would consist of modules for: 1) segmenting the spine; 2) modeling spinal components; 3) detecting, assessing and analyzing salient features; and 4) generating information for medical diagnosis, patient reporting and clinical management. Despite a clear need for a versatile spine assessment toolkit, none is commercially available today. Some components of the proposed toolkit have been researched by our group in the past, but have neither been fully developed for cross-platform application nor made robust enough for clinical use. The major novelty of this project is creating the application software. Other novelties include adapting algorithms and code so that the component modules handle CT and MR images, deriving a set of validation metrics for a cross-platform application, designing the shell program, developing the test protocols, creating a feature set and distilling information that could be used for patient management and clinical decision-making.

Bin Gu
Faculty Supervisor: 
Dr. Mousumi Bhaduri