Multiple sclerosis (MS) is a neurodegenerative disease that mainly affects the brain. Defining your diagnosis and monitoring treatment is of vital importance. Due to neuronal damage, which occurs in neurons, this disease affects the ability to communicate nerve cells in the brain and spinal cord. The causes of this disease still present a gap. Magnetic resonance imaging (MRI), due to its richness in the detail of information provided, is the gold standard examination for diagnosis and follow-up of multiple sclerosis. MS is characterized by brain lesions where the neurodegeneration process occurs, and it is possible to visualize these affected areas in magnetic resonance imaging. The advancement of the technology has allowed an improvement in the visual detection sequences of lesions caused by multiple sclerosis, aiding the medical diagnosis, but there is still a need for an objective quantification of the injured area. The use of computer programs to evaluate sclerotic lesions in MRI has been a differential to promote a more objective quantification, to determine the pathological extension and to follow the progression of the disease. These computational techniques can be developed through digital imaging studies, which can be applied to various types of medical images. Therefore, the proposal of this project is the development of a computational tool based on the deep neural network method for the detection and quantification of brain lesions in patients with multiple sclerosis.
News published in Agência FAPESP Newsletter about the scholarship: