Dados do Trabalho

Título

Radiomics in the study of spinal cord injuries caused by Multiple Sclerosis, MNOSD and HTLV-I virus infection.

Resumo

Introduction: Radiomics is an emerging field that uses Artificial Intelligence for medical research. Radiological imaging exams can be statistically analyzed in a way that allows extracting multiple graphical elements not visible in routine assessments and transforming them into numerical data. It is a method that extracts a large number of features from medical images using data-characterization algorithms. The main steps of radiomics analysis include image collection as well as reconstruction, segmentation of the region of interest (ROI), feature extraction as well as quantification, and establishment of the predictive as well as prognostic models. Through this technology, it is possible to analyze the heterogeneity of tissues and lesions at the cellular level, helping to distinguish pathophysiological processes in a non-invasive way. Compared to biopsy studies, in radiomics, the whole tissue or the whole lesion is available for analysis, and not just a short sample. In this way, radiomics plays an important role in daily practice, once it can distinguish diseases that eventually look alike radiologically, but hold completely different pathophysiology and treatments. In our case, the diseases under study are MS, NMOSD, and HTLV-I virus, a common infectious disease in our country. Objectives: Our objective is to use the technology of radiomics to analyze, in a non-invasive way, spinal cord injuries caused by Multiple Sclerosis, MNOSD and HTLV-I virus infection. Thus, it is possible to identify the structure and components of the lesions, helping to distinguish the different medical conditions and guiding to adequate treatment, as well as accompanying the structural differences during treatment or surveillance. The need for a consistent number of patients for the study brings to us a necessary inclusion of patients from different Centers in Brazil, thus comprising a multicenter study. Methods: This is an ongoing study in which patients enrolled are those in whom there is a spinal lesion with common characteristics for MS, NMOSD, or HTLV-1 disease and who will be submitted under radiological investigation with the aid of radiomics technology to differentiate in a non-invasive and unequivocal way these three conditions. Conclusions: This study will allow us to demonstrate the great potential of radiomics in the diagnosis and monitoring of neurological diseases. We invite neurologists from other centers to take place in this project.