ИСТИНА

Abstract. Identifying imaging biomarkers (IBs) of stroke remains a priority in neurodiagnostics. There is a number of different methods for image analysis and learning rules applicable in this field, but all of them require large arrays of DICOM images and clinical data. In order to amass such dataset, we have designed a platform for systematic collection of clinical data and medical images in different modalities. The platform provides easy-to-use tools to create formalized radiology reports, contour and tag the regions of interest (ROIs) on the DICOM images, and extract radiomics data. Subsequent analysis of the obtained data will allow identifying the most relevant IBs that predict clinical outcome and possible complications. The results of the analysis will be used to develop predictive algorithms for stroke diagnostics.