Scope
When surgical planning with a software tool, the problem of tissue with diffuse edges segmentation is very often faced. That is the case when physicians try to segment soft abdominal tissues.
This problem is coped in this project with the development of semiautomatic segmentation algorithm that, instead of being based exclusively in the Hounsfield value of each voxel, employs additional information such as shape and regularity of edges.
In other words, minimization of an energy function is the criterion for segmentation. This energy function is penalized when edges are abrupt, with a lot of inlet and protrusions, or when the final shape of the segmented object is not the desired. Likewise, similarly to traditional segmentation algorithms, this energy function decreases when the difference between Hounsfield values within the object and in the outer border is high.
There exist different techniques to find the correct contour by minimizing this energy function. In this project level-sets, graph-cut and convex relaxation techniques are being tested.
For this purpose the segmentation of retroperitoneal tumors has been addressed and the validation data has been made publicly available and can be downloaded (see download link below).
The segmentation of bone and muscle structures has been also addressed and the corresponding validation data has also been made publicly available. It can be downloaded from the download link below.
Collaborations
Fundación Pública Andaluza para la Gestión de la Investigación en Salud de Sevilla (FISEVI)
Hospitales Universitarios Virgen del Rocío
Funding
Segmentación de órganos abdominales en imágenes TAC, FISEVI, Junta de Andalucía.
Images
Publications
- Mendoza, C.S., Acha, B., Serrano, C. Linearized multi-dimensional earth mover’s distance gradient flows. Submitted to IEEE Transactions on Image Processing.
- Mendoza, C.S., Bohórquez-Ruiz, G., Acha, B., Serrano, C. Multidimensional earth movers distance active contours. Submitted to International Conference on Image Processing, ICIP 2011.
- Suárez-Mejías C, Pérez-Carrasco JA, Serrano C, López-Guerra JL, Parra-Calderón C, Gómez-Cía T, Acha B (2016) Three-dimensional segmentation of retroperitoneal masses using continuous convex relaxation and accumulated gradient distance for radiotherapy planning. Medical & Biological Engineering & Computing. DOI: 10.1007/s11517-016-1505-x.
- Suárez-Mejías C, Pérez-Carrasco JA, Serrano C, López-Guerra JL, Gómez-Cía T, Parra-Calderón C, Acha B (2016) Validation of a method for Retroperitoneal Tumor Segmentation. International Journal of Computer Assisted Radiology and Surgery. In Review. Download USevillatumorDATABASE.
- Pérez-Carrasco JA, Acha B , Suárez-Mejías C, López-Guerra, J.L., and Serrano Carmen. Joint Segmentation of Bones and Muscles Using an Intensity and Histogram-Based Energy Minimization Approach. Submitted to Computer Methods and Programs in Biomedicine. Download Database at USevillaBoneandMuscleDATABASE.
