Program Information
IBEX: An Open Software Infrastructure Platform to Accelerate Collaborative Work On Quantitative Image Features and Predictive Models
L Zhang1*, L Hunter2 , D Fried1 , X Fave1 , J Yang1 , L Court1 , (1) MD Anderson Cancer Center, Houston, TX, (2) Baylor College of Medicine, Houston, TX
Presentations
SU-D-9A-5 Sunday 2:05PM - 3:00PM Room: 9APurpose: Quantitative imaging analysis has great potential to improve the accuracy and efficiency of cancer research, but the lack of software infrastructure prohibits progress from developing and using its methods. We have developed a software platform to flexibly support common quantitative imaging analysis tasks such as multimodality image data import and review, feature algorithm development, predictive model design, and consistent data sharing among multi-institutions.
Methods: The IBEX(Imaging Biomarker Explorer) software package was developed under Matlab and C/C++. The software architecture deploys the modern Model-View-Controller (MVC) and unit testing programming concepts to isolate each task and to validate if their relevant data and algorithms are fit for use. The View component represents multimodality images with delineated structures (if available), which are imported using DICOM, CERR, or other native formats. Non-image data, such as patient performance status can also be entered. The Controller component represents image pre-processing and feature extraction algorithms. The Model component represents the predictive model formula and coefficients. Thanks to the unit testing implementation, the user has the option of reviewing the corresponding result to check the quality of data and algorithms at each stage. While IBEX itself is self-contained and has implemented typical algorithms and models, it is an open system and additional algorithms and models can be easily added by defining them in library files. Also, a complete external model can be imported, including necessary data about pre-processing algorithms, feature algorithms, model formula, and model coefficients. This greatly helps maintain data consistency when institutions attempt to validate external models.
Results: Testers found the IBEX software to be intuitive and easy to use. All import, pre-processing algorithms, and feature algorithms were fully validated.
Conclusion: We have successfully implemented an open software infrastructure platform to accelerate work on quantitative imaging analysis and collaborative research activities.
Funding Support, Disclosures, and Conflict of Interest: This work was funded by the NCI(1R03CA178495-01)
Contact Email: