Improved 3-D Delineation Algorithm for use with Multiple Medical Imaging Applications

Technology #13240

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Engin Dikici
Richard D. White M.D.
Managed By
Zahara M. Jaffer
Assistant Director 352-392-8929
Patent Protection
US Patent 9,177,373


The University of Florida is seeking companies interested in applying and commercializing a new and improved algorithm for processing three-dimensional images used in many types of medical diagnosis. Medical imaging is widely used for many purposes, including locating and evaluating tumors and diagnosis of various diseases. This novel approach developed by researchers at the University of Florida to estimate volume of objects (e.g. tumors, organs) imaged by 3-D scans greatly reduces time needed to analyze the object of interest. This practical algorithm is intuitive to use, allowing easy integration with a Graphic User Interface (GUI) and any imaging tool or Picture Archiving and Communication Systems (PACS). The major advantage is much reduced time needed working with scanned images to generate an outline of the object of interest for volume estimation.


For use in medical image processing systems


  • Reduced need for manual operator inputs to generate volume estimates used for diagnosis, thus improving efficiency
  • Permits easy integration with any imaging tool and PACS program, offering practicality, intuitive use, and minimizing adaptation costs
  • Offers modality- and anatomy-independent image processing, eliminating need for adjustments and allowing multiple market applications


This technology allows integrated use across all imaging applications, providing greater flexibility and expanded applications. Because it uses unique polar thin plate splines – a novel algorithm – instead of active contour models, the technology is anatomy independent, and does not require adjustments when finding surfaces for different objects. It can be used for any anatomy of any approximately spherical surface shape without any investigations prior to use. The new algorithm eliminates the need for the point set used to be a subset of predefined landmark positions, and eliminates the need for a knowledge base of surfaces, previously extracted from other studies, in order to find a new surface. This technology also allows real-time adjustment of the extracted surface, without requiring image processing.