IN129-ED-THB7

Cancer and Phenomics Toolkit (CAPTk): A Software Suite for Computational Oncology and Radiomics

Thursday, Dec. 1 12:45PM - 1:15PM Room: IN Community, Learning Center Station #7



FDA Discussions may include off-label uses.

Participants
Sarthak Pati, Philadelphia, PA (Presenter) Nothing to Disclose
Saima Rathore, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Ratheesh Kalarot, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Patmaa Sridharan, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Mark Bergman, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Taki Shinohara, PhD, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Paul Yushkevich, PhD, Philadelphia, PA (Abstract Co-Author) Investigator, KinetiCor, Inc
Yong Fan, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Ragini Verma, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Despina Kontos, PhD, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose
Christos Davatzikos, PhD, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose

CONCLUSION

CAPTk provides a bridge for quick translation between academic research and clinical application, thereby enabling translation of cutting-edge research into clinically practical tools quickly and efficiently.

FIGURE

http://abstract.rsna.org/uploads/2016/16014589/16014589_uw28.jpg

Background

Availability of highly sophisticated methods that help us gain a comprehensive understanding of the underlying mechanisms of cancer, reveal substantive insight into the biological basis of disease susceptibility, treatment response, and identification of new drug targets, has skyrocketed. Translating such academic research into clinical practice is one of the biggest challenges, due to their complexity. We present CAPTk as a tool that facilitates translation of such algorithms from medical imaging research to the clinic. It replicates basic functionality of radiological workstations (accepts MRI data in DICOM & NIfTI formats with manual annotations). It is distributed under a completely free BSD-style license and is designed to be computationally efficient.

Evaluation

CAPTk has been designed for advanced image analysis pertaining to oncological conditions in the brain, breast and lung. It provides a springboard for using a number of algorithms, including but not limited to image registration, thresholding, bias correction, quantitative image analysis applications for predicting the status of Epidermal Growth Factor Receptor variant III in glioblastoma (GBM), semi-automatic mask generation of anatomical structures, evaluation of the peritumoral spatial heterogeneity in GBM, as well as analysis of normal tissues for predicting risk of future cancer. Utilizing industrially validated & open source tools (ITK, VTK, Qt), CAPTk encapsulates the core principles of good software design (usability, stability, scalability), supports all major deployment platforms and provides interfaces to help developers integrate C++ & Python applications utilizing a common interface.

Discussion

Users will see an easy-to-use graphical interface to access cutting-edge methods, along with image annotations (drawing points, tumor radius, masks), and developers can easily add applications to CAPTk while having access to the interactive & visualization routines. Examples will be showcased during the demonstration.