Advanced search
Start date

Content-based retrieval of medical images to aid the clinical decision using radiomics

Grant number: 17/23780-2
Support type:Scholarships in Brazil - Doctorate
Effective date (Start): August 01, 2018
Effective date (End): March 31, 2021
Field of knowledge:Physical Sciences and Mathematics - Computer Science - Computational Mathematics
Principal Investigator:Agma Juci Machado Traina
Grantee:Jonathan da Silva Ramos
Home Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
Associated research grant:16/17078-0 - Mining, indexing and visualizing Big Data in clinical decision support systems (MIVisBD), AP.TEM


Nowadays there is a huge amount of images daily produced. There is no difference in the medical scenario, in which a lot of images are produced in many exams and need to be stored and analyzed. However, the manual analysis and annotation is a time consuming task and requires a great knowledge for each image analyzed. To aid in this task, content-based image retrieval - CBIR systems are widely used. Notwithstanding, CBIR systems are mostly isolated systems that seeks just one output (anomaly or no anomaly present). To improve even more CBIR systems, the radiomics systems were proposed, which aims at extracting as many features as possible from the images and combine them with other clinical data. In the end, those information can be mined and analyzed, improving the individual treatment and patient tracking. On the other hand, there are many challenges to implement a radiomics system. This doctoral project aims at advancing in the CBIR and CAD fields by integrating the radiomics system, seeking to maximize to the highest level the security and confidence of those systems. Initially, we are working with spinal images, which presents a considerable challenge when there is deformation or lesions present. The project's contributions will be analyzed by specialists in the Hospital das Clínicas in the medicine faculty in Ribeirão Preto - HCFMRP, within the thematic FAPESP project MIVisBD. (AU)