Analytics for computation and visualization of liver resections. (ALIVE)
Planning liver surgical resections is an extremely complex task. In this process, surgeons decide the trajectory separating the healthy part of the liver from the part that will be resected. Although liver resection has been practiced for decades, among surgerons, there is still no consensus on what a good resection is. An underlying problem contributing to this lack of consensus is the fact that there are no formal methods to specify and communicate resection plsns-clinicians often use subjective descriptions (written or oral) or hand-drawings. In ALive, we believe that liver surgical practice needs computer-assisted tools allowing the surgeons to specify resection plans in a flexible way, thus accommodating different surgery planning techniques and cultures. We also believe that providing a way to represent resections in a formal way can contribute to a better evaluation and communication of resection plans. ALive is a research project funded by the IKTPluss program from The Research Council of Norway.
ALive aims to improve liver surgery practice by researching new methods for planning and visualization of liver resections. The project will work on two different areas:
- Generation of resection plans using geometric modeling, AI and vasculature analysis.
- Formal representation and visualization of 3D resections in lower dimensions (2D).
UCO is responsible of the application of High Performance Computing techniques in Liver Resection Planning and Liver Analytics and Visualization.
Liver resection planning is the only stage in the state-of-the-art processing pipeline that has not been subject to automation. Obtaining an algorithm that can suggest resections not only can reduce the time invested in surgical planning, but also can be the only feasible planning option for surgical planning in complex cases (e.g., multiple metastases). In addition, the possibility of applying AI from a hybrid pool of data (vascular structures, medical images, segmentations and 3D models) can provide deeper understanding of AI processes.
Generating parameterized vascular territories can greatly contribute to the different interpretations needed for different clinical applications-as suggested by Fasel and Schenk. This has implications not only for the clinical practice, but also for the clinical research, where parameterized vascular territories can be a tool for its advancement.
We believe that obtaining formal representations of resections in a compact form that is interpretable by clinical experts has very large implications for liver surgical practice. These advancements will contribute to the documentation of real cases, for quality assurance, education and communication (e.g., science and medical records).
All the novelties considered in this project have a great potential to reach the industry as part of existing medical platforms. Our resectograms proposal could even become an industry standard for the documentation and communication of liver resections.
Informática Avanzada (GIIA)
Code PAIDI: TIC-252
Joaquín Olivares Bueno. Partner.
Universidad de Córdoba
Budget of Andalusian group: € 23,717.35
