20 Jul DEEP LEARNING BASED STEREOTACTIC CRANIAL SURGERY PLANNING
To improve the life expectancy and quality of a patient, early diagnosis, medical attention and accurate analysis are required. In this work, we propose a highly scalable system, with a focus on generalizability to other domains, that is capable of end-to-end cranial surgery planning, being the first study to define the surgery planning operation as an optimization problem and solving it via deep learning. The system distills the knowledge of the doctors and creates ensembles that will theoretically plan better cranial surgeries. To achieve this, four state-of-the art models: tumor classification, tumor segmentation, atlas-based segmentation, and tractography and a novel algorithm specifically designed to represent, calculate and minimize the risks involved therein are proposed.
Category: COMPUTING — Country: TURKEY — Year: 2021
EMİRHAN KURTULUŞ
Video:
Poster:
